Landslide is the main geological disaster that threatens the safe operation of natural gas pipelines. It is of great significance to investigate the analysis method that is suitable for the evolution law of landslide and the mechanical response of overburden pipes in engineering. Therefore, based on the principle of the minimum potential energy, the total potential energy balance equation of slope system (from anti-sliding section to sliding section), i.e., the Lagrange variational equation, was firstly adopted to obtain the critical conditions for the instability and slip of the slope. When the strain energy of the sliding section is equal to the strain energy for the crack and penetration of the anti-sliding section, the slope will slip. Meanwhile, the strain energy stored in the slope will be transformed into kinetic energy. On one hand, according to the contact relationship between the rock mass and the pipeline during the sliding process, the mechanical response model of the pipeline in two landslide stages is proposed. On the other hand, considering the fragmentation of the rock and soil mass, the uniform stress model of pipeline under the action of rock landslide and the non-uniform stress model of soil slope pipeline are proposed respectively. In view of this, based on the mechanical response for the small-scale interaction of the pipeline and soil, the stress expression of the pipeline in the pre- and post-sliding stages of rock and soil slopes, which is suitable in the elastic part, is derived. Finally, taking the EES244 section of natural gas pipeline from Sichuan to East as the research target, the total potential energy equation of the slope system is established, the situations for deformation, instability and slip of the slope are analyzed, the stress values of the pipeline in different sliding stages are calculated, and the safety of the pipeline is evaluated. Meanwhile, the stress analysis and the safety verification of the pipeline are carried out from a large-scale perspective by means of the numerical method. The results show that a sudden change of stress in the pipeline will be occurred near the interface between the sliding zone and the non-sliding zone, and the internal stress in the sliding zone increases slightly, while the whole pipeline is in a safe and stable state. Therefore, combining the small-scale theoretical calculation and the large-scale numerical method is of great significance and practical value, especially for the preliminary designing of the proposed pipeline, the safety evaluation of the existing pipeline and the later maintenance of the pipeline.

The rice husk ash, which is environmental friendly waste-recycling, can strengthen the soil. Triaxial tests were conducted to study the stress-strain properties, strength characteristics, and variation characteristics at different strain levels including the elastic modulus, deviator stress and reinforcement strength ratio of rice husk ash mixed clay with different proportions of rice husk ash and its reinforced soil. The test results show that the maximum dry density of the mixed soil reduces significantly and the optimum moisture obviously increases with the increase of rice husk ash content. The content of rice hush ash has a great influence on the shear strength of the reinforced soil, and rice husk ash-soil mixture’ s initial tangent modulus and peak stress that addition of 10%?15% rice husk ash is recommended as the maximum. Compared with geotextile reinforced rice-husk-ash mixed soil, higher deviatoric stress and shear strength are observed on specimens with geogrid, and this trend is more obvious with the increase of the number of reinforcement layers. It also shows that the inflection point of the stress-strain curve is more prominent with the increase of the number of reinforcement layers. The elastic modulus of the mixture material depends on the content of rice husk ash, the type and layers of reinforcement. After adding the rice husk ash, the elastic modulus of the soil increases significantly, and it increases 1.5 times in better content when reinforced by geogrid, both rice husk ash and the reinforcement can effectively strengthen the soils. As the number of reinforced layer increases, the reinforcement strength of rice husk ash-soil mixture increases significantly but has little relation with the confining pressure.

With the aid of damage mechanics and on the basis of statistical strength theory, a method for establishing the constitutive model of deep fractured rock mass is proposed and verified by laboratory and numerical tests. The fractured rock mass is divided into numerous micro-cubes. The strength of micro-cubes is related to the fracture degree and the strength of each micro-cube is randomly distributed. Thus the strength can be used to reflect the fracture degree of the rock mass. Among them, based on the fact that the work done by friction between fracture surfaces is equal to the strain energy released after the material fracture, the rock fracture degree variable defined from the mechanical point of view is obtained. In addition, it is assumed that the strength distribution of micro-cubes obeys the Weibull distribution and the stress behavior satisfies Hoek-Brown criterion. The constitutive model of fractured argillaceous sandstone rock mass is then established and verified based on typical triaxial test results of fractured rock samples. The results show that the calculated curve from theoretical model is in good agreement with the test results. At last, a supplementary numerical test is carried out using discrete element software PFC, which further proves the good calculation performance of theoretical model for argillaceous sandstone and the feasibility of the constitutive model establishing method proposed in this paper.

Based on the large shaking table test of saturated sand on the free field, the dynamic shear stress–strain response of the foundation soil model was obtained by using the linear interpolation method according to the data from accelerometer arrays. At the same time, the concepts of apparent viscosity and zero shear viscosity of non-Newtonian fluid based on the assumption of hydrodynamic theory were introduced to study the characteristics and behaviors of shear thinning after site liquefaction. The characteristics of solid-liquid phase transition of saturated sand under seismic load were studied. Results showed that when the saturated sand was stimulated by a large earthquake, excess pore water pressure accumulated rapidly, and sand liquefaction occurred as the pore water pressure ratio of the upper part of the saturated soil ran up to 1. According to the variation of dynamic shear stress–strain curve, the dynamic shear modulus of saturated sand decreased significantly after liquefaction, indicating that the soil softened gradually. The variation trend of shear stress-shear strain rate of saturated sand derived from the data of dynamic shear stress and dynamic shear strain was similar to the rheological curve of non-Newtonian fluid. The apparent viscosity of the liquefied soil in the upper part of saturated sand layer decreased significantly. After liquefaction, the sand showed the pseudoplastic fluid characteristics of "shear thinning".

The phenomenon of silting and spalling is very common during glutenite weathering in Maijishan grottoes. Previous studies have shown that the participation of salt greatly accelerates the weathering process of rock. In this paper, several sets of cylindrical specimens made of glutenite were used in two tests, namely capillary migration test and cycle degradation test. The migration patterns of sulfate in samples and the failure characteristics of the glutenite specimens were discussed. Based on the experimental data and theoretical analysis, the crystallization pressure in glutenite samples was calculated. This is an important parameter for judging the damage of glutenite caused by salt crystallization. The results show that: the crystallization of sodium sulfate caused substantial damage to glutenite; the migration of sulfate in glutenite had regularity; when the concentration of sodium sulfate solution was between 0.95 mol/L and 1.13 mol/L, the destruction of the conglomerate began. The theoretical maximum crystallization pressure could reach 33 MPa under the experimental conditions.

Laterite is very sensitive to water. This drawback can be substantially mitigated by adding alkaline materials like lime, but adding alkaline materials can impair the long-term performance of laterite due to the weak acidity of laterite. In this study, a certain amount of lime (5%) and metakaolin (4%) (La+L+MK) was added into the laterite to improve its water sensitivity and mitigate the deteriorative acid-base reaction. The mixtures with eight different initial water contents were compacted and cured at predetermined periods. After the curing, unconfined compressive strength (UCS), calcium ion concentration, electro conductibility and pH were tested. The results show that the UCS of the stabilized soil approached its peak at the water content of 26%, so it was not conducive to the strength development of the stabilized soil when the water content diverged from the optimal value. The reason is that when lime is not well hydrated due to lack of water, calcium ion cannot be released, thereby inhibiting the pozzolanic reaction. As a result, cementitious hydrates in the inter-aggregates of laterite cannot be generated. Also, the bonding strength increase due to the pozzolanic reaction is lower than the matrix suction loss caused by excessive moisture when the sample is too wet. The measurements of calcium ion concentration and electro conductivity confirmed the above conjectures. Our results clearly show that metakaolin combined with lime significantly improve the laterite strength at the wet state. Even after being saturated, the relative stability of the lime-stabilized soil can still be maintained with metakaolin additives, which indicates that metakaolin can effectively reduce the water sensitivity of lime soil and improve its durability. This phenomenon is due to the fact that metakaolin contains amorphous silicon and aluminum oxides and has edge-surface-contacted structures at the microscopic scale, thereby enabling it to significantly reduce the soil pH to the alkaline range that favors dissolution of silica and alumina oxides, thus accelerating the pozzolanic reaction and slowing down or inhibiting the deterioration reaction.

Taking the Beijing-Shanghai high-speed railway CRH380BL EMU as the prototype, a 1:10 ratio railway ballast track subgrade shaking table model test was carried out based on the law of similarity to analyze the acceleration, earth pressure and displacement response of the subgrade. The model is 9.6 m×3.5 m×1.0 m (length×width×height), including the train, ballasted track and subgrade part. Some findings are as follow. As the altitude increases, the amplification factor of peak horizontal acceleration increases gradually, and the value is basically stable between 1.0 and 2.5; while the amplification factor of peak vertical acceleration firstly increases and then decreases, and the value is basically stable within 1.5. As the input ground motion intensity increases, the amplification factor of the peak horizontal acceleration is directly proportional to the elevation, and the non-linear relationship is gradually strengthened. The maximum amplification factor of vertical peak acceleration is shifted from the bottom H/3 to 2H/3 and the peak magnification of horizontal and vertical accelerations reach the maximum when the input seismic wave PGA is 0.3g. As the input ground motion intensity increases, the peak earth pressure intensity in the filler increases gradually. When the input seismic wave PGA reaches 0.4g, the earth pressure intensity reaches the maximum. The earth pressure intensity at the center section of the subgrade tends to increase firstly and then decrease with the increase of elevation, and the maximum earth pressure gradually shifts from H/3 to 2H/3 at the bottom of the subgrade. When the input seismic wave PGA is 0.05g, the earth pressure intensity on the subgrade surface and the subgrade bottom is linearly distributed along the horizontal direction of the subgrade. The farther the former is from the subgrade, the greater the earth pressure, and the latter is basically unchanged. When the input seismic wave PGA is 0.15g, 0.30g and 0.40g, the earth pressure intensity is the smallest at the edge of the ballast, followed by the earth pressure intensity at the center of the subgrade, and the earth pressure presents a triangular distribution along the horizontal direction of the subgrade; the horizontal displacement of the middle and top of the subgrade slope gradually increases, and the former is smaller than the latter, showing an approximately linear distribution and the gradual reduction in the difference between them. At the top of the subgrade, the difference between the displacement on the slope and that on the top of the center line of the subgrade increases gradually with the intensity of the input ground motion intensity, and the difference between both gradually increases, and the stable finally; When the seismic wave is at the bottom of the embankment, the main frequency is concentrated in 5?15 Hz. As the elevation increases, the subgrade has a strong amplification effect on the 30?40 Hz frequency band, but the influence on the remaining frequency bands is not significant.

This study aims to develop a nonlinear constitutive model to address the mechanical properties of magnesia-cement- solidified soil under different contents of solidified agent (Wg) on the basis of the results of a series of laboratory tests. Firstly, a new type of environmentally magnesia cement composite curing agent was adopted to solidify the sludge. Then, the scanning electron microscope (SEM) tests, the one-dimensional consolidation tests and the undrained triaxial tests were carried out on the magnesia cement solidified soil with different contents of Wg. The test results show that there exists a strong structure in the inner of the magnesia-cement-solidified soil and the associated structural yield stress becomes larger and larger with the increase of Wg. The cohesion between the soil particles increased due to the binding material produced by the hydration reaction, which produces the material to fill the soil void. As such, the stress-strain behaviour of the magnesia-cement-solidified soil transformed from work hardening to work softening. Based on the aforementioned test results, a constitutive model was developed to investigate the transformation of the stress-strain relationship of the magnesia-cement-solidified soil, in which the effect of solidified agent content (Wg) could be also accounted for. The numerical examples show that the proposed constitutive model behaves well in predicting the stress-strain behaviour of the magnesia-cement-solidified soil at any amount of Wg.

This research investigated the relationship between temperature (25?900 °C) and the microstructure and mechanical properties of sandstone through the application of nuclear magnetic resonance, electron microscopy, X-ray diffraction and uniaxial compression techniques. The result firstly suggested temperature has a significant impact on sandstone microstructure because an increase in temperature can improve the total porosity. This phenomenon is especially obvious when temperature is higher than 600 °C, in which the high temperature quickly increases the number of medium and large sized pores, which in turn lead to a significant permeability increase. Meanwhile, a temperature rise may also lead to a decrease in elastic modulus, an increase in peak strain and an elongation in pore compaction stage, these can be reflected by a decrease in brittleness and an obvious enhancement in plasticity. In addition, the mechanical strength of sandstone under heat condition may also be impacted by mineral composition. Such impact was not significant when temperature was below 450 °C. That is because low temperature has minor impact on mineral composition, in this way, the rock strength was mainly influenced by porosity. But when temperature raised to 450?600 °C, the dehydration of kaolinite and the formation of new phase minerals increased the rock strength dramatically, this offset the effects of porosity increase, therefore increased the strength of sandstone. However, when temperature exceeds 600 °C, the rock strength decreases again because of a rapid increase in large sized pores.

Moisture migration inside unsaturated silty subgrades under dynamic loading is of great significance for investigating the stability of heavy haul railway subgrades in water immersion conditions. Based on soil electrical properties, a measurement system combined with a GDS cyclic triaxial instrument was developed relying on the van der Pauw principle to measure the water content distribution properties of unsaturated soil specimens. With the use of the aforementioned measurement system, water-supply dynamic triaxial tests were carried out. Moreover, the stratified water content of the unsaturated specimen was continuously measured during the test to explore the influence of the initial compaction degree of the unsaturated specimen and dynamic stress amplitude of dynamic loading on water migration inside the unsaturated specimen. This measurement system which can be used to determine the stratified water content of unsaturated specimens in real time, is able to achieve nondestructive and consecutive testing during dynamic triaxial tests, and the maximum error is only 0.7%. As the water content of the unsaturated soil specimen gradually increases under dynamic loading, the influence of the soil specimen initial compaction degree on resistivity decreases, and the influence can be ignored when the water content is close to saturated water content. A function exists to correlate soil resistivity with soil specimen initial compaction degree and water content, and the function has a high level of correlation and uniqueness. According to the analysis of the testing results, under the combined action of rainfall and cyclic loading, the water inside the subgrade accumulates to a certain depth inside the soil, and the pore water pressure in this area increases. The increase of pore water pressure decreases the strength and leads to the occurrence of mud-pumping disease. Increasing the subgrade compaction degree and the axle weight of the train will inhibit the downward water migration inside the subgrade and contribute to subgrade stability. However, an excessive increase in axle load may result in subgrade instability. When the axle load is greater than the critical dynamic stress, the subgrade will be destroyed.

Under seismic loading, the pile foundations in the liquefiable soil are often destroyed due to the liquefaction of foundation soil. In this process, even if the soil does not reach full liquefaction finally, the strength of saturated sand will be weakened due to the existence of the excess pore water pressure, which will also lead to the decrease of horizontal resistance force of soil. If the influence of excess pore water pressure on horizontal resistance force of soil is not considered, the pile foundations are still designed by adopting the p-y curves in the API standard, and the results will be more dangerous. In this paper, the dynamic cyclic torsional shear tests are carried out for the Fujian standard sands by employing the vertical-torsional coupling shear apparatus, and the dynamic characteristics and weakened parameters of saturated sands in different weakened states are studied. Then the formulas of ultimate soil resistance are derived based on the improved theoretical model of soil wedge at shallow layer. Combined with the theoretical model of flow failure around piles at deep layer, the ultimate soil resistances at different pore pressure ratios at any depth are obtained, and then the p-y curves of pile-soil interaction in saturated sand foundation in different weakened states are constructed. It can be found from the study that the pore pressure ratio, which characterizes the weakened state of soil, has a significant effect on the ultimate soil resistance in pile-soil interaction. With the increase of pore pressure ratio, the weakened degree of soil will be more serious, and the ultimate resistance of saturated sand is smaller. That is to say, the action of the lateral loaded pile on the surrounding soil decreases with the increase of soil weakened degree, and the vice versa.

In order to study the bearing behavior of a flexible single pile under combined horizontal and uplift loading, two groups of structural testing, comprised of 6 pile models in total have been carried out. The results show that for a flexible single pile, pre-applied uplift load with a magnitude smaller than half of pure uplifting capacity(Tu1) lead to significant improvement of ultimate horizontal bearing capacity, achieving greater horizontal resistance compared to pure ultimate horizontal bearing capacity (Hu1) Additionally, it has been observed that this horizontal bearing capacity increases initially but decreases subsequently under further increasing uplift load. This load increasing results in reduction of bending moment and soil resistance whilst degrades horizontal stiffness, and consequently, the horizontal displacement of pile increases initially but decreases subsequently under resultant effect. The ultimate uplift bearing capacity of a flexible single pile could be improved under pre-applied horizontal loads not exceeding 0.5Hu1, mainly attributed by increasing the average side friction of pile within the distance of 10 times the diameter of pile below the ground surface. This capacity is strengthened under larger horizontal load, and when this load reaches 0.5Hu1, the magnitude is increased by 17.1% compared with Tu1. This behavior could be taken into account with respect to pile design. The precision of test results has been verified by theoretical analysis, and a formula for calculating the ultimate uplift bearing capacity of flexible single pile under combined loads has been proposed. The corresponding analysis results have achieved good agreements with testing data and the error is less than 6%.

In order to explore the strengthening mechanism that the ultimate bearing capacity of single pile with cap is greater than the sum of the ultimate bearing capacities of single pile and cap foundation under the synergistic action of pile-soil-cap, two-dimensional discrete element models (DEMs) which can reflect the processes of static load test of single pile with cap, single pile only and cap foundation only in sand are established, and the rationality of the models are verified by comparing the model-predicted results with the variation laws of load-settlement curve in the laboratory model test. Based on the validated DEMs, the deformation mechanism of the sand foundation and the bearing characteristics of pile foundation are analyzed from the microscopic perspective. The results show that, in terms of the sand foundation deformation mechanism, compared with single pile and cap foundation, the foundation soil is more significantly disturbed by the single pile with cap, and the soil in a large range under the cap participates in the process of resisting the load. As for the bearing characteristics of the pile foundation, the compactness of the soil near the pile side and pile end is strengthened by the effect of the arch formation that occurs around the pile side of the single pile with cap, which improves the lateral friction resistance and the bearing capacity of pile end. As a result, the overall bearing capacity of the single pile with cap is improved. The results of DEM can provide references for the design of pile foundation with cap and the bearing capacity estimation method.

To maintain the stability of excavation face in slurry shield, a filter cake must be timely formed during the excavation work. With the filter cake, the slurry pressure can counteract the earth pressure and water pressure in the excavation stratum. In this study, a mathematical model of slurry penetration under constant pressure in sand stratum is established by considering the filtration effect. By means of this model, the influence of penetration time, slurry concentration, slurry pressure and initial stratum porosity on soil porosity and filter cake formation is analyzed. Combining the periodic disturbance of cutting-tools, the variation of soil porosity on the excavation face caused by multiple slurry penetration is summarized. Finally, the supporting mechanism of the excavation face is discussed during the shield tunneling based on the design of cutting wheel and tunneling parameters. The study results show that the slurry penetration behaviour is mainly affected by the characteristics of slurry stratum without disturbance of cutting-tools. In contrast, with the periodic disturbance of cutting-tools, the decreasing cycle of cutting will increase the difficulty of filter cake formation under the same cutting depth. In addition, the filter cake formation and the stability of excavation face can be significantly improved if the number of cutting-tools on same cutting track is reduced and the advance speed of shield and rotate speed of cutting wheel are decreased appropriately.

To clarify the sliding failure mechanism of the shield tunnel face and determine the reasonable range of the supporting force during shield construction, a three-dimensional slip rupture model for the shield tunnel face was proposed by using the spatial discretization technique based on the slip line theory and the upper bound theorem of limit analysis. According to the large principal stress arch theory, the vertical earth pressure at the top of the slip zone was calculated, and the limit supporting force of the tunnel was calculated by using the vertical load on the upper part of the slip damage zone. The results show that the soil arching effect significantly affects the magnitude and the distribution of vertical stress in front of the tunnel face. By comparing proposed model with existing approaches, it is found that the upper bound solution of limit supporting force obtained from the new model has good applicability in both purely cohesive soils and frictional soils. At the same time, the shape of the damaged area on the tunnel face is fairly consistent with the results from the centrifuge model test and numerical calculation.

The filling and bonding effects of hydrate increase the compactness and strength for gas hydrate-bearing sediments(GHBS), which makes the GHBS exhibiting properties similar to dense sand or cemented soil. Under the frame of unified hardening model of clay and sand (CSUH model), the mechanical properties of GHBS are summarized firstly, and a compressive hardening parameter is introduced to describe the isotropic compression characteristics of GHBS under the double influences of filling and bonding of hydrate. A bonding parameter is put forward to modify the yield function, and an evolution rule of bonding effect is also proposed. The state parameters are used to adjust the dilatancy equation to reflect the dilatancy and softening depending on density. Thus, an elastoplastic model is developed, which can describe the strength, stiffness, shear dilation and strain-softening of GHBS. The model coded and tested, and the simulation results are compared with the experimental ones of GHBS. The results show that the proposed model can well describe the stress-strain relationship, shear contraction with hardening and shear dilation with softening for GHBS.

Expansive soil is a widely distributed soil with obvious dilatancy, shrinkage and crack-rich properties. Its crack initiation and propagation are affected by many factors. In this study, the image processing and crack quantitative analysis program developed by MATLAB were used to explore the size and temperature effects of the crack evolution of expansive soil samples. Firstly, twenty-five expansive soil mud samples with different initial states were prepared. Then, the evolution of cracks on the surface and the corresponding water content of the samples during the shrinkage due to water loss were recorded using a designed camera system. Finally, using crack quantitative analysis program, some indexes including fracture degree, aspect ratio, average crack width and fractal dimension, were quantitatively analyzed under the size effect or temperature effect, and the size-temperature combined action, respectively. It is shown that the fractal dimension value of the crack remains stable, and is only slightly affected by the thickness of the samples. But there exists an approximate logarithmic relationship between the fractal dimension and the moisture content. In the process of water loss and shrinkage, the final fracture index of the expansive soil was mainly affected by the thickness of the sample, and the surface size also had a certain impact on the final value of the fracture length to diameter ratio and width. Temperature accelerates fracture development and stability. However, the impact of average width value and temperature on the final index was not obvious. The thickness was the most significant factor, followed by the surface size, and finally temperature. Additionally, the microcosmic mechanism of shrinkage and cracking process of expansive soil were also analyzed according to this experiment.

Brazilian disk splitting tests were conducted on the granite disk specimens to study the feasibility of indexes based on AE parameters (i.e., RA and AF) for illustrating the fracturing intensity of rock failure. The distributions of RA and AF during the loading process were studied and compared with the results of SiGMA moment tensor analysis. The results show that the number of the shear cracks increases with approaching the damage stage even in the tensile failure of rock. The suggested judgement value for analyzing the failure mechanism based on RA and AF was proposed. An index r=RA/AF was proposed to illustrate the fracturing intensity of rock failure and was further compared with the amplitude parameter of distribution, variation and attenuation characteristics. The results show that the evaluation based on the index r leads to more conservative results. Furthermore, the coefficient of variation of r was proposed to be the statistical index of r and compared with b value, which is the statistical index of the amplitude. The results verified the feasibility of CV(r) in evaluating rock failure and the judgement value of granite splitting failure was suggested according to the b value.

Reinforced cushion has been widely used in engineering structures such as highways and railway subgrades subjected to repeated loading and unloading due to its many advantages. To study the deformation characteristics of reinforced cushion under loading and unloading, a single loading and unloading static test was carried out for the two models of gabion mesh reinforced and unreinforced cushions. Variations of vertical deformation of the cushion and reinforcement strain under different loads were monitored and obtained. The deformation characteristics of the reinforced and unreinforced cushions were compared and analyzed, and the working mechanism of the reinforced cushion was discussed in terms of energy. The results show that the settlement and residual deformations of the reinforced cushion (at the location of the loading plate) under loading are smaller than those without reinforcement. The vertical deformation of the surface of the cushion after reinforcement (outside the range of the loading plate) and its affected range have increased. The shape of the horizontal distribution curve of the vertical deformation during the unloading process changes from concave to convex, and its loading and unloading curve shows a shape of “∞”. The reinforcement strain shows a non-uniform distribution along the transverse direction, and the maximum strain is less than 0.06%. The reinforcements are always in an elastic deformation state. The reinforcement in the reinforced cushion has the effect of storing, releasing and laterally transmitting strain energy, which results in better bearing capacity and elastic property of the reinforced cushion, thereby reducing plastic deformation or cumulative deformation of the cushion under cyclic loading and unloading.

In order to provide reference for the seismic design of the portal in the bridge-tunnel connected system, the failure process of the portal was studied through the large-scale shaking table tests, and obtained the acceleration signals which were quantitatively analyzed by wavelet packet transform. It is found that the failure of the portal is marked with the appearance of the crack in the vault slope. The unsafety degree of tunnel portal in the bridge-tunnel connected system ranks as the vault, the abutment, the slope and the invert in turns under the action of ground motion. The strength and stiffness of the bridge and tunnel are relatively high and therefore, the failure often starts from the slope, then the vault and abutment, while the damage of the invert is generally slight. The seismic wave with a low-frequency (0.1?12.51 Hz) plays a leading role in the failure process of the portal-slope system. Due to the complex refraction and reflection effect between the different material interfaces, the low-frequency waves will change drastically before propagating from soil into concrete structures. By analyzing the changes in the percentage of the energy of the low-frequency waves, the failure of portal-slope system can be divided into three stages, i.e., the elastic deformation stage under the small-shake, the small elastic-plastic deformation stage under the medium-shake, and the large failure deformation stage under strong-shake.

Irregular columnar joints network models were firstly prepared by using the Voronoi diagram random simulation method and 3D printing technology. White cement slurries were chosen as the rock-like material and poured into the 3D printing irregular columnar joints network models. Then the single columns were removed from the 3D printing models, and the white latex and 502 glue were used to bond the single columns together. In doing this, the irregular columnar jointed rock mass specimens were obtained. Laboratory uniaxial compression tests were carried out on these specimens to investigate the strength and failure modes. The experimental results show that the changes in the uniaxial compressive strength against the dip angle β resembled a "J" shape trend, which indicates that the anisotropy of the uniaxial compressive strength is remarkable. Three typical failure modes are summarized, including splitting failure along the columnar joints, shear failure along the columnar joints and the axial compression failure along the single columns, and the failure patterns are somewhat different from those of the regular columnar jointed rock mass specimens. Furthermore, the experimental results in this study were compared with previous results with respect to the columnar jointed rock mass specimens bonded by pure cement slurry. The results show that the weaken effect of the columnar joints can be more adequately reflected using the specimens bonded by white latex and 502 glue than those bonded by cement slurry.

The modified Poulos method is universally used in design of laterally-loaded pile groups of platforms in current ocean engineering practices. However, given that increasing engineering cases have jack-up rigs with large-diameter spudcan foundations operating near a platform, the effects of spudcan penetration on the adjacent pile group should be evaluated. A pile group analysis method which is able to evaluate the effect of spudcan penetration on the pile group interaction is proposed in this paper. Single pile head deflections under the combined loading of spudcan penetration and the pile head load are firstly analyzed using the nonlinear foundation beam method. Then combining the effects of pile group interaction and nonlinear soil-pile interaction, the additional pile response due to the group effect is calculated through an elastic analysis. Finally, the pile group interaction Y-modifier is obtained by several pilot modifications of p-y curves of the single pile, whose variation during spudcan penetration indicates the effect of the spudcan penetration on the pile group interaction. Centrifuge model tests are conducted in sand to study the effect of spudcan penetration on adjacent a single pile and a pile group, respectively, and the proposed method is verified by comparison with predicted pile head deflections at different stages. In this case, the Y-modifier tends to decrease during spudcan penetration, and it can be inferred that the factor determined in the active loading stage is conservative enough in pile group analysis when considering spudcan penetration.

Rubber-sand mixture, as a kind of low-cost and environmental protection geotechnical material, has broad applications, and the research on its mechanical properties is of great significance. In this study, the combined test of shear wave and compression wave for rubber-sand mixture is carried out by bender-extender element method. The propagation time of shear wave and compression wave is determined by discrete frequency scanning method and initial arrival wave method, respectively. After that, the elastic dynamic parameters such as initial shear modulus, constrained modulus and Poisson's ratio, are obtained. The influence of rubber content and confining pressure on elastic dynamic parameters of rubber-sand mixture is analyzed as well. The results show that under the same confining pressure, with the increase of rubber content, the initial shear modulus and constrained modulus of rubber-sand mixture gradually decrease, and the initial Poisson's ratio gradually increases. Under the same rubber content, with the increase of confining pressure, the initial shear modulus and constrained modulus of rubber-sand mixture gradually increase, and the initial Poisson's ratio gradually decreases. Finally, the coupling effect of the two factors and the related mechanical mechanism are analyzed.

Based on the triple-shear failure criterion of materials and the mechanical properties of unsaturated soils, the triple-shear failure criteria of the single stress variable and the double stress variable for unsaturated soils are put forward, and their characteristics are analyzed. The results show that the existing failure criteria for unsaturated soils can be interpreted approximately with the new criteria by changing the influence coefficient b of the principal stress. For loci on the π plane, the failure criteria proposed in this paper cover all the convex regions from the inner boundary of the single-shear failure criterion to the outer boundary of the triple-shear failure criterion. Therefore, the new criteria are suitable for unsaturated soils under various complex stress states and can reflect the inequality in uniaxial tension and compression of the unsaturated soils. In addition, several true triaxial test results reported by other researchers are adopted to verify the proposed criteria. For the unsaturated clayey sands, the true triaxial test results were found in good agreement with the values predicted by the triple-shear failure criteria of the single stress variable and the double stress variable when the influence coefficient of the principal stress b=0.6, and the predicted values given by the double stress variable criterion are more consistent than that of the single stress variable criterion. For the unsaturated loess, true triaxial test values are also in good agreement with the values predicted by the triple-shear failure criteria when the influence coefficient of principal stress b=0.2, and the difference is smaller between the two kinds of triple-failure failure criteria than that of the clayey sands.

In order to study the dynamic response characteristics of pile foundation under the pile-soil-fault coupling effect, four different types of seismic waves with the peak acceleration of 0.35g are selected for shaking table test. The acceleration response, relative displacement of pile top, bending moment of pile body and damage of pile foundation under different types of seismic waves are studied. The test results show that the parameters of the pile on hanging wall are obviously larger than those on the footwall of the fault, showing the hanging wall effect. The peak acceleration of pile top is greater than that of pile bottom, and the upper soil layer has filtering effect on the input seismic wave. The acceleration response of pile top has hysteresis compared with that of pile bottom. The acceleration at the top of the pile and the amplification coefficient α of the peak acceleration are the largest for the El-Centro wave. The α difference between the hanging wall and footwall is the largest for the Kobe wave. The peak values of relative displacement of pile top and bending moment of pile foundation are the largest for the Kobe wave. The bending moment of pile body is larger at the soil interface. When different types of seismic wave are applied, the peak value of bending moment of pile body does not exceed the bending capacity of pile body. It is suggested that the pile foundation of bridge near fault in strong earthquake area can be checked according to different seismic waveforms.

The equivalent response acceleration method for transverse seismic analysis of underground structures is systematically studied in this paper. Firstly, results of the dynamic time history analysis method are set as the benchmark to evaluate the adaptability of the equivalent response acceleration method through numerical examples. Then, error sources of the equivalent response acceleration method are dissected and overcome. An improved equivalent response acceleration method is proposed by introducing the dynamic response adjustment coefficient β. Finally, the applicability of the proposed method is evaluated through numerical examples. Results show that, the results of internal forces and deformations obtained from the equivalent response acceleration method are all smaller than those from dynamic time history analysis method. Moreover, its accuracy can be improved with the increase of soil shear velocity but less affected by the buried depth. Further research finds that the error of the equivalent response acceleration method comes from the fact that, the interference effect of the underground structure on the dynamic response of the surrounding soil is ignored in the method. After improvement and optimization, the improved equivalent response acceleration method is proposed to further consider the interference effect. Compared to the traditional equivalent response acceleration method, the proposed method has a wider applicability and better calculation accuracy, which provides a novel approach for seismic design and analysis of underground structures.

The breakdown pressure of the pressure-time curve obtained from the in-situ test is an important parameter for calculating the tectonic stress. In order to investigate the influence of pumping rate on breakdown pressure and pressurization rate, four large-scale hydraulic fracturing experiments on low-permeability hard and brittle limestone with varying constant pumping rate were conducted. Acoustic emission monitoring technique was used to investigate how varying pumping rate affected breakdown pressure, failure mode and complexity of fracture network as well as internal relationship between pumping rate and pressurization rate. The experimental results show that the pumping rate has a significant effect on the breakdown pressure. A larger pumping rate results in a higher breakdown pressure and a lower complexity of fracture network. The typical pressure-time curve consists of a slow boosting stage, a rapid boosting stage, a stable boosting stage and a sudden drop stage. Pressure increases linearly with time during the stable boosting stage, and there is a clear linear relationship between the pressurization rate and the pumping rate. Using the linear relationship between the pumping rate and the steady-state pressurization rate, the Ito theory can quantitatively explain the dependence of the breakdown pressure on the pumping rate. The theoretical predictions are in good agreement with the experimental results.

It has been popular and difficult for scientists and engineers to acquire more microseismic information, automatically identify microseismic data of rock rupture, locate the location of rock rupture automatically and accurately, and provide theoretical and technical support for disaster analysis, early warning, and prevention and control. In view of these difficulties, an integrated and high precision intelligence microseismic monitoring technology was developed. In the microseismic monitoring technology, several innovation techniques are developed, which include that a sensing-acquisition-transmission technology was integrated; the acquisition technology with noise reduction of 32 bits A/D coupled with electric components was established; integrated identification method of microseismic signal including recursive STA/LTA method and BP neural network method was proposed; the microseismic source location algorithm based on the technique of PTP high-precision time synchronization and the algorithm of velocity model fast matching in database was developed. At present, this technique has been applied in many deep rock engineering at home and abroad. The application results show that the developed technique can be applied not only to monitoring dynamic disasters such as rock burst, but also to monitoring the stability of surrounding rockmass in excavation (mining) of rock engineering, and also to monitoring anti-theft mining of mineral resources. From these application cases, it has been show that the capacity of acquiring microseismic signals, identifing microseismic signals of rock rupture automatically, and locating the location of rock rupture accurately was improved during rock engineering disaster evolution using the developed microseismic technique. The developed technique can promote the rapid development of microseismic monitoring technology towards automatic monitoring, analysis and intelligent early warning of rock engineering disasters.

A method for calculating the unloading additional stress in the soil at the bottom of deep foundation pits under the condition of layered excavation is proposed, based on the Mindlin solution. The nonlinear relationship between the soil resilient modulus and the unloading additional stress during foundation pit excavation is considered in the calculation of rebound deformation using the layered summation method. The calculation results show that affected by the depth of layered unloading, the unloading additional stress under the condition of layered excavation is less than that under the condition of one-time excavation within the most of the depth range, and the difference between them is obvious within three times the depth of foundation pit below the pit bottom. With the same foundation pit area, the unloading additional stress at the bottom of long strip foundation pit is less than that of square foundation pit. This method can be used to evaluate the influence of pit bottom reinforcement on the unloading additional stress of soil and rebound deformation combining with the equivalent layer method. With the increase of resilient modulus or thickness of reinforced layer, the unloading additional stress and the rebound deformation decrease, and the extent of deformation reduction decreases correspondingly. The method has been verified to be reasonable through the case study of West Square of Hangzhou East Railway Station. Based on the parameter analysis, the equivalent design method of resilient modulus and thickness of the reinforced layer is proposed, which has guiding significance for practical engineering.

In order to characterize the transportation mechanism of the high-speed rockslide, it is significant to investigate the tribological characteristics of the sliding surface. However, the investigation of tribological characteristics has not been thoroughly revealed. Taking the Wangshan-Zhuakousi rockslide in Emeishan city of Sichuan province as instance, the engineering geological conditions and the motion characteristics of the rockslide were obtained through field geological survey and video surveillance. Based on the high-speed friction experiments of the rockslide masses (basalt) and the sliding plane (tuff) by the pin and disc worn instrument, the tribological characteristics (dynamic friction coefficient and microscopic wear surface appearance) were analyzed under dry and saturated conditions. The results show that: the dynamic friction coefficient of the sliding plane (the interface between basalt and tuff) is much smaller than that of basalts. Regardless of the dry and saturated conditions, the dynamic friction coefficient of the sliding plane (the interface between basalt and tuff) is negatively correlated with both the frictional velocities and the normal pressures in the experiments. Regardless of the dry and saturated conditions, the dynamic friction coefficient of sliding mass (basalt) is also negatively correlated with the frictional velocities, whereas it is positively correlated with the normal pressure in the experiments. In the high-speed friction process, the dynamic friction coefficient is increased instantly due to the dilative shear failure of the sliding surface. The experimental results can be adopted for explaining the high-speed motion of the hard rockslides containing the weak interlayers. In addition, the findings in this study can provide essential design parameters for disaster prevention and high-speed rockslide mitigation in the tuff and basalt areas.

The spatial effect of the foundation pit on the tunnel has a great influence on the tunnel deformation. In this paper, a new model of unloading ratio considering spatial effect is developed. It can comprehensively consider the influences of the depth，width and longitudinal length along the tunnel of the upper foundation pit. The three-dimensional finite element method and field measurement method are employed to determine the influence of spatial effect of the upper foundation pit on the tunnel deformation. Numerical result and field monitoring data show that the deformation of the shield tunnel is mainly caused by upper excavation unloading in the main overburden area over the shield tunnel, the construction of foundation pits outside the main overburden area has little effect on the maximum uplift deformation of tunnel, and in the main overburden area the maximum uplift deformation of tunnel is approximately linear with the unloading ratio proposed in this paper. In soft soil area the uplift deformation of the tunnel by upper excavation unloading can be controlled by controlling the unloading ratio.

Dynamic stability evaluation and failure mechanism research were performed on the Yanyang village landslide along Xiangli expressway in Yunnan province. The landslide stability analysis under different seismic intensities was carried out, the progressive failure of landslide was described by the change in volume ratio of residual elastic zone of slip zone. The dynamic stability of landslide was evaluated by combining with the deformation mode of landslide and the volume ratio of residual elastic zone. For the failure mechanism of landslide under extreme earthquake conditions, the failure process of landslide was described in terms of time and space respectively. A cusp catastrophe model of landslide, which could consider both weakening and hardening section of slip zone, was established and the trigger mechanism was revealed. The results showed that: (1) the landslide under the condition of Ⅷ degree seismic intensity remained stable, and only local failure occurred due to the "locking effect " of locking section; (2) The critical peak acceleration of overall failure of the landslide was 2.29 m/s2, and its failure mechanism was the whole failure caused by the sudden penetration of the plastic zone due to the failure of "locking action" under the coupling action of leading section traction and trailing section tension crack; (3) The leading, middle and trailing section of the slide zone were not destroyed synchronously, but presented a cumulative-triggering process. (4) A failure criterion of stiffness effect was derived based on the improved cusp catastrophe model, the overall stability of the landslide was found to be closely related to the stiffness and size characteristics of the sliding zone medium. The results could offer guidance for disaster prevention and seismic design of Yanyang village landslide, and be used for reference in the dynamic stability evaluation and failure mechanism analysis of similar projects.

To investigate the stability of the surrounding rock during excavation process of the underground caverns at the Baihetan hydropower station, effects of excavation unloading on the deformation and failure of the surrounding rock using the numerical simulation software 3DEC based on DEM (discrete element method) are analyzed. The microseismic monitoring technology is introduced to monitor and analyze micro-fracture evolution inside the surrounding rock of the caverns in real time, and the numerical simulation results are compared with the microseismic monitoring data. In order to verify the accuracy of numerical simulation and microseismic monitoring results, the spatiotemporal evolution law of the macro-deformation of the surrounding rock is studied by conventional displacement monitoring. The results show that the damage of surrounding rock is closely related to the on-site construction state and also affected by various geological structures. The deformation characteristics of the surrounding rock obtained by numerical simulation are basically consistent with the aggregation rules of microseismic events obtained by microseismic monitoring, and the results are in good agreement with the conventional monitoring results. The comprehensive research method combining the three-dimensional discrete element numerical simulation and microseismic monitoring technology can better describe the mechanical behavior of the surrounding rock under excavation unloading, and effectively evaluate the damage characteristics and potential risk areas of surrounding rock.

In order to simulate the transition of rock from continua to discontinua more accurately and reduce the mesh dependence, an element splitting method is developed based on the continuum-discontinuum method where the Lagrangian element method is coupled with discrete element method. The computational domain is discretized into high-precision quadrilateral elements; after cracking, the cracks can propagate along the diagonal lines of quadrilateral elements and the edges of triangular or quadrilateral elements. Deformation-cracking processes of the Brazilian disc rock specimen, uniaxial compression rock specimen and surrounding rocks under hydrostatic pressure are simulated. The following results are found. (1) For the Brazilian disc rock specimen, tensile cracks propagate from the center of the disc to the top and bottom until they penetrate through the disc, and the tensile cracks are relatively smooth; for the uniaxial compression rock specimen, shear cracks are relatively straight, and the direction of the main shear crack penetrating through the rock specimen is consistent with the shear plane direction obtained according to the Mohr-Coulomb criterion. (2) From the effects of hydrostatic pressure and internal friction angle on deformation-cracking processes, the following results are found. Under hydrostatic pressure, firstly, V-shaped notches appear near the cavern surface due to shear crack propagation; then, long and curved shear cracks appear due to further shear crack propagation whose distribution is similar to shear slip-lines. With an increase of hydrostatic pressure, the ranges of shear cracks increase, whereas with an increase of internal friction angle, the ranges of shear cracks decrease, and the angle between the propagation direction of the long shear crack and the annular direction decreases. (3) According to the simulation of the surrounding rock in phase-Ⅱ Tianshengqiao hydropower station, four V-shaped notches are observed, which is in a good agreement with field observation.

The reliable recognition of strata in front of tunnel face is significant for the stability and safety of the tunnel engineering project. Traditional advanced geological forecasting methods could not ensure high identification accuracy, low cost and short construction time simultaneously, and they can’t satisfy the universality of stratum identification under different geological conditions. The advanced forecasting efficiency could be significantly enhanced if the drilling data of surrounding rocks in front of the tunnel face can be obtained while performing the conventional advanced borehole to attain the rock conditions at different drilling depths in real time, which would be convenient and efficient by not affecting the construction period. However, no objective and accurate stratum identification methods are found. In this paper, we proposed an intelligence analysis of drilling data and stratum recognition method based on neural network. It is used to analyze the advanced drilling test data of Jiudingshan Tunnel of Chuxiong-Dali highway and the analysis method was verified by the strata exposed after tunnel excavation. The results show that the error rate of stratum recognition using the single drilling parameter is about 35%. The combination of blow energy and blow number, water pressure and water flow water cannot significantly improve the accuracy of stratum recognition. The combination of drilling speed, torque, rotation speed and propulsion can reduce the error rate to 22% for stratum recognition. The error rate can be sharply decreased by 9%~12% when the standard deviation of drilling parameters is introduced into the neural network model. The error rate of stratum recognition is less than 10% for random sampled data and it is less than 14% for a single borehole using the neural network model with the combination of multiple drilling test parameters.

It is a way to measure the internal force of pile foundation by sliding micrometer, and the anti-seepage measures during the installation of sliding micrometer tube are the key to achieve high quality test, while quantitative research on anti-seepage effect of measures is rare. A set of device for testing the anti-seepage ability of sliding micrometer tube was developed. The seepage flow of three common anti-seepage measures under different pressure differences was tested, and a formula for calculating the seepage flow was developed. The matters needing attention in the installation of the sliding micrometer tube were analyzed according to the site conditions. Test results show that when the internal pressure is higher than the external pressure, the seepage flow rate is much larger than that when the external pressure is larger than the internal pressure. Glue coating is a very effective anti-seepage measure, and generally no particles with a diameter over 0.005 mm will enter the sliding micrometer tube after adopting this measure. Increasing the water head in the sliding micrometer tube is beneficial to reduce impurities entering it, but the water head in the sliding micrometer tube can not be too much higher than the liquid levels in the borehole of the pile. After pouring concrete, if water overflow in the sliding micrometer tube happens, the orifice of the sliding micrometer tube should be closed to limit water overflow, and if the water level in the sliding micrometer tube falls, the orifice of the sliding micrometer tube should be opened to allow gas to enter it.