The rock strength will decrease under a certain stress, which can lead to different performances in engineering such as the abrupt destroying by strength weakening, long-term strength and so on. So the research on time dependence of rock strength is very important to construction safety and long-term stability of rockmass engineering. Based on the analysis of a large number of test results, evolution characteristics of rock strength and their mechanisms are summarized, and it is can be found that strength weakening of rock is related to the inherent strength of rock, the stress state (yield approach index) and the performing duration of stress. Therefore, time-dependent theoretical model of rock strength is proposed by properly theoretically characterizing the evolutions of cohesion and frictional angle of rock. Simulations to some sets of test data show that the proposed model can perfectly and quantitatively describe the evolution of weakening process of rock strength; moreover, the model is of clear physical concept. The effort of this paper gives a new method to study the time-dependence of rock strength.

To validate the stabilization effect of stabilizing agent GX08 treating the marine and lacustrine soft soil in Hangzhou, and to investigate the adverse influence of organic matter on cementing soft soil, unconfined compressive strength (UCS) tests are performed on stabilized soils. Test results indicate that organic matter significantly hinder the growth of the strength of stabilized soil, whereas stabilizing agent GX08 can effectively enhance the strength of the stabilized soil. The results show that the strength of stabilized soil is well correlated with organic matter content as a form of quadratic function, and it is correlated linearly with cement content, and both are correlated as the form of logarithmic function with stabilizing agent GX08 content and curing time. The total cement/water is applied to the establishment of the stabilized soil strength model. Based on the analysis of tests data, a predict model, considering the influences of organic matter content, cement content, stabilizing agent content and curing time on the strength of stabilized soil, is proposed to predict strength development. Finally, the model is to be promoted to use, so as to verify the applicability of the model.

Shear and tensile strengths are two aspects of the mechanical properties of structural soil. The strength of the structural intact loess and remolded loess are analyzed through two kinds of tests, which are namely the tensile strength test and the conventional triaxial shear strength test. The hyperbola strength formula of structural intact loess and remolded loess are determined by combining the test results of tensile and shear strength. The algorithms of tensile-failure stress state, tensile-shear stress state and compression-shear stress state are further investigated by modifying the corresponding failure stress according to the nonlinear strength. The structural parameter of stress ratio and the structure property are introduced into the developed hyperbola strength theory of structural loess. The results show that the established hyperbola strength formula of structured loess effectively overcomes the defect of exaggerated tensile strength based on the Mohr-Coulomb strength theory. It provides a practical strength theory which can evaluate the engineering disasters in structural loess area.

An uniaxial creep test of frozen red sandstone is conducted under -10℃, and stress levels are 0.36, 0.45 and 0.54 separately. The result shows that the red sandstone creep is stable when the stress levels are 0.36 and 0.45, but the rock creep is unstable when the stress level is 0.54. The higher the stress level is, the longer it will take for the creep reaches stability. From the analysis of derivative function of test data, it is found that the threshold value of acceleration stage in unstable creep is the 13rd hour. Three empirical equations of the red sandstone creep are gained by fitting the test data. A creep model is established which installs Kelvin body, modified Maxwell body and modified CYJ body in series and gains the parameters of the model. The model curves are consistent with test data better; so the model reveals creep discipline of frozen red sandstone. The result of creep test provides some valuable references for the evaluation of frozen rock’s stability in the mines that are being constructed by freezing method.

The damage failure mechanism and mechanical characteristics of jointed rock mass have been studied through freezing- thawing cycles and similar material test method. Through the observation of the damage failure modes and uniaxial compression test of the samples after freezing-thawing cycles, the effects of joint dip angle, joint continuity degree, joint set, joint filling depth, sort of joint filling, sample’s saturation degree and number of cycles on the damage failure mode, uniaixal compression strength and elastic modulus are studied. It is shown that the existence of joint and its physico-mechanical properties have much effect on the damage failure modes and strength. The joint dip angle affects the failure modes and strength by affecting the occurrence position of freezing-thawing cracks. The cracks become more with joint continuity degree. The effect of freezing-thawing on the samples becomes worse with number of joints. The freezing-thawing damage degree first increases then decreases with joint filling depth. The sort of joint filling has certain effect on the freezing-thawing damage degree. The freezing-thawing damage degree first decreases then increases with the sample’s saturation degree. The cracks become more and wider with the number of cycles; and its anti-freezing-thawing property is worse than that of the intact one. The above research conclusions can provide valuable references to the constructions and safety operation of rock mass projects in cold regions.

The hydraulic fracturing of rock mass cracking is one of the important influencing factors that causes rock mass cracking and leakage and even construction water gushing, which is also the research hotspot in geotechnical engineering field. The disturbance effect on crack propagation of aquiferous rock cracking by blasting excavation is analyzed based on fracture mechanics; and the results show that, under disturbance of blasting excavation, the cracking propagation of aquiferous rock relates to the strength and incident angle of blasting stress wave, the value and direction of ground stress, the value of pore water pressure, the crack dip, and the fracture toughness, etc. The effect of blasting stress wave equals to increasing the pore water pressure inside rock mass cracks, which means per 1 cm/s peak particle vibration velocity equals to increasing 0.1 MPa pore water pressure; and the larger the blasting vibration velocity is, the greater the blasting disturbance load will be. The change of ground stress and pore water pressure caused by rock mass excavation around rock mass cracks have complex influence on instability and propagation of cracks, which may change the instability and propagation model of cracks.

The physico-mechanical properties of tailings dam material will be changed by capillary water, which will effect on the stability of tailings dam. The research on this field is barely reported at home and abroad. Therefore, in-situ sampling and laboratory geotechnical experiments on tailings dam material are conducted for testing its physical properties, and the rising test of capillary water is carried out about 6 months by the self-designed layered sampling device for capillary fringe; and the physico-mechanical properties of tailings sand in capillary water fringe are gained by layered sampling and soil testing. Furthermore, the variation characteristics of physico-mechanical indexes such as water content, gravity density, cohesion, internal friction angle and shear strength with the height of capillary water have been systematically studied. The results show that water content and gravity density decreased with the height of capillary water. The relationships of cohesion and internal friction angle with the height and water content increases firstly and then decreases. However, the relationship of internal friction angle with water content is different from the traditional conclusion that the relationship is decreasing with water content. This difference is caused by surface properties of the tailings particles which are different from that of sands. The surface of tailings particles is dirtier than that of sands and is covered by mineral and something else which is soluble in water. The research results also show that the effect of capillarity on shear strength is great, and the relationship of shear strength with the capillary height increases firstly and then decreases. As we known, the gravity is one of major causes of tailings dam failure. In a real tailings pond, there is full of industrial wastewater which has bigger surface tension than pure water, and the bigger surface tension will cause higher capillary water in tailings dam. In fact, the capillary water will result in weight gain of tailings dam which is danger to the stability of tailings dam. Therefore, it is more important to analyzing the stability of tailings dam by researching on the physico-mechanical properties of dam body materials in capillary water fringe of tailings dam. This research will put forward a new idea to the dam and slope engineering, and laid a foundation for establishing a new method and theory to analyzing stability of tailings dam by considering the role of capillarity.

Rheological phenomenon of the concrete faced rockfill dam can be critical according to prototype monitoring. In view of the three-dimensional shear rheological rate formulas of rockfill material are different in different documents. Three-dimensional rheological rate of rockfill material is studied with the rheological model of 3 parameters as an example. Volume rheology and shear rheology are assumed different. First of all, the three-dimensional volume rheological rate formula is deduced; and then, based on Prandtl-Reuss flow rule, it can be derived by Mises yield function as a plastic potential function. The computational formula of three-dimensional shear rheological rate of rockfill material is derived taking associated flow rule and assuming that the plastic potential function is Mises yield function. The theory of the computational formula of three-dimensional rheological rate is rigorous, since the computational formula of three-dimensional rheological rate can degrade into the computational formula of uniaxial state. Thus it is pointed out that the theory of the computational formula of three-dimensional rheological rate in related papers is not rigorous.

Through the indoor model tests, lateral earth pressures exerted by natural foundation and rigid pile composite foundation against rigid retaining walls without any displacements are measured. The distribution of lateral earth pressures of rigid pile composite foundation is obtained by the comparison with that of natural foundation. It is believed that participation of pile in rigid pile composite foundation makes magnitude and distribution of lateral earth pressure different from natural foundation. In this place, pile participation includes the deep transfer function of pile load, the influence of pile negative friction and the pile "block" effect on level additional stress of soil between piles. The influence range of lateral earth pressure of rigid pile composite foundation with surcharge is deeper than that of natural foundation. Besides, the value of lateral earth pressure of rigid pile composite foundation is smaller than that of natural foundation at the same location under the same subgrade pressure. In the test conditions of this study, the total and the additional earth pressure exerted by the rigid pile composite foundation on the rigid retaining walls are about 43.3%~80.1% and 15.9%~59.8% of the total earth pressure of the natural foundation respectively when the load level reaches the characteristic value of the foundation capacity within the range of 0.35 to1.4 m from the rigid retaining wall to the building.

Freeze-thaw test and triaxial compression test are carried out for model samples with pre-existing single fracture by the rock mechanical servo-controlled testing system, freeze-thaw damage degradation model is studied based on the freeze-thaw test, and crack coalescence mechanism is discussed under the triaxial compression condition. Tests reveal that: freeze-thaw damage deterioration modes of fractured rock masses are scattered to particle model, crack model and prefabricated crack fracture mode. Under the triaxial loading condition, there are four coalescence modes of fractured rock samples due to freeze-thaw cycles: tensile mode，shear mode，compression mode and mix mode, which are closely related to the freeze-thaw cycles, confining pressure and the crack dip angle. With the increase of freeze-thaw cycles and confining pressure, more and more surface ruptures are found, at the same time, crack coalescence mode has experienced from single to mix; the failure modes of rock samples show tensile and compression mode when confining pressure reaches the values of 2 MPa and 6 MPa, while tensile mode are presented when the confining pressure reaches the value of 4 MPa. Rock samples with smaller fissure angle which equals 30° are taken on tensile mode, however, the rock samples with the longer fissure angle of 60° take on shear mode.

In South China, the geosynthetic reinforced retaining wall using high liquid limit soil as backfill is applied widely. Because of climate change, the suction between unsaturated high liquid limit clay and geotextile is altering; thus influencing the interface shear characteristic. So the interface shear characteristic between unsaturated high clay and geotextile considering suction change should be studied deeply. In this paper, a series of direct shear tests are performed using ZFY-1A direct shear apparatus to investigate the shearing resistance properties of high plasticity clay-geotextile interface under suction cycle. Results of experiment indicate that the peak shear strength of the soil-geotextile interface increases with the net normal stress. Interface strength increases at first and then reduces with the increase of cycling times, interface strength is lowest through 3 times of the dry-wet circulation. And interface strength is reduced in general. , , , and increase and then decrease with the increase of cycling times. increases with the cycling times and net normal stress. decreases with net normal stress, and increases then decreases with the increase of cycling times. Finally, the interface strength formula under different suction conditions is obtained through the logarithmic fitting.

Moraine soil is a kind of special geotechnical material generated by rapid mixed depositing of boulders, gravel, sand, silt and clay without sorting when quaternary glacier moving and melting. Engineering problems caused by moraine soil have attracted more and more attentions from geotechnical engineering field. Meso-fabric of gravel particles is a reflection of the unique sedimentary history and directly affects the geotechnical properties of moraine soil. However, subjecting to the hardness of obtaining undisturbed moraine soil samples, there is few researches on it. To change this situation, we succeeded in getting undisturbed moraine soil samples from west Yulong Snow Mountain with delicate manual cutting work. Then the 6 pieces of moraine soil samples are scanned by CT instrument and the real meso-fabric images of internal moraine soil are obtained for the first time. After binarization of CT images and de-noising processing with the help of Matlab, section fabric indexes of stone area ratio, grain orientation and particle shape of moraine soil gravel particles are given with conventional statistical methods. The results show that the difference between vertical sections is minor, however the difference between vertical and horizontal sections is significant.

Soil arching effect in geogrid reinforced pile supported embankment under train dynamic load is studied by 3D model tests. Based on the test results, it is found that the soil arching effect remains valid under dynamic loading and the soil arching effect will be enhanced when the fill height increases and the existing geogrid will also enhance the soil arching effect. However, the dynamic stress distribution in geogrid reinforced pile supported embankment under dynamic loading is entirely different from the stress distribution under dead weight. The soil arching will degenerate under dynamic loading, which manifests that a part of dynamic stress previously carried by piles will transfer to soil between piles during dynamic loading applied and the dead load initially carried by piles will also transfer to soil between piles after finishing dynamic load application. The fill height and geogrid have large influences on soil arching under dynamic load; but when the fill height reaches a certain height, the influence of fill height on soil arching will be negligible; the influence of dynamic load on soil arching is weakened by half by geogrid.

An explicit analytical solution of the partial differential equations of consolidation by vertical drains is derived to assess the effects of compressibility and varying hydraulic conductivities of soils in the smear zone on ground consolidation. Equal volumetric strain assumption is adopted. The compressive deformation of soils in the smear zone is considered, together with a linear and a parabolic distribution of the horizontal hydraulic conductivity along the radial direction. Well resistance is also considered, together with an arbitrarily distributed mean stress increasing along the depth of ground. The solution is derived for the consolidation under a linearly time-varied surcharge preloading. The proposed solution is used to analyze the influences of the radius of smear zone, the distribution of horizontal hydraulic conductivity and the coefficient of volumetric compressibility of smeared soils on the global average degree of consolidation. It is found that the rate of ground consolidation is underestimated by the solution derived by applying a uniformly reduced horizontal hydraulic conductivity to the smeared soils. When the radius of smear zone is relatively large, the rate of ground consolidation is overestimated by the solution derived without consideration of the compressive deformation of the smeared soils. These effects should not be overlooked in the analysis of consolidation by vertical drains.

The pile foundation of super-highrise buildings is generally installed into the deep sandy strata of confined aquifer. Deep excavation is generally conducted for the basement over the aquifer. In order to keep the stability of pit bottom, dewatering is required to conduct in the confined aquifer. Since there are a large number of piles in the dewatering pit in aquifer, the existing piles will affect the path of groundwater seepage. In order to reveal the mechanism of blocking effect of piles on groundwater seepage, laboratory tests are conducted. Polyvinylchloride (PVC) pipes and sand are employed to simulate pile and test soil, respectively. Test results show that piles show an obvious blocking effect on groundwater seepage; meanwhile the quantity and the layout form of piles will affect the blocking effect of piles. The effective media theory can be used to calculate the coefficient of permeability by considering the piles with rectangular layout. However, there is a difference of 20% with the quincunx layout, owing to the prolonged flow path. The formula to calculate the modified equivalent hydraulic conductivity is established based on the analysis of the flow path.

In order to solve the inclined problem of the ancient seawall, the earth pressure on the ancient seawall with different types of backfills are observed based on the in-situ test. Three types of ancient seawall are filled with different backfills as the foamed cement banking (FCB), natural soil and shale ceramisite respectively. The law that the earth pressure changed with time is analyzed, also including the distribution of the earth pressure and the location of resultant force point. The unloading effects of the three different filled materials are discussed. The distribution of earth pressure is nonlinearly folded at different levels in the middle of the wall. The earth pressure is higher in the middle and lower at the both top and bottom of ancient seawall. The height of actual action point of earth pressure resultant from field measuring is 0.42-0.49 times of the wall height which is upper than the theoretic calculating result. FCB is an effective unloading filled material used to decrease the pressure on the ancient seawall that it can reduce the earth pressure and overturning moment by 33.1% and 41.4% respectively comparing with undisturbed soil.

The ice cementation effects to the structure surface bond soil particles and infrastructure together during soil freezing. This cementing force is named freezing strength between frozen soil and foundation which is usually measured by means of the shearing strength of the interface. Hence, the strength characteristics as well as the relations between shearing stress and shearing strain of interface are very important to analyze the interaction between structures and the permafrost and to determine the uplift performance of foundation in permafrost regions. For better engineering application, a series of laboratory direct shear tests on the adfreezing interface between frozen soil and concrete plate are performed to investigate the mechanical behavior of the interface. Based on the stress and shear displacement relation of the interface and standard modeling method, a stress-displacement-temperature constitutive equation of the interface is deduced to describe the mechanical behaviors of the adfreezing interface. Experimental results indicate that the constitutive model can reasonably describe the stress-displacement principles for different temperatures. The constitutive relation can also be used in the numerical simulation of stress and deformation of the structures in cold regions.

Soft soil generally has creep characteristics, the creep on microstructure shows the changes of soil particles and pores. In order to understand internal mechanism of the long-term deformation, a consolidated undrained triaxial creep test about undisturbed and remolded of the Tianjin Binhai dredger fill is carried out. At the same time, multi-group parallel tests under the confining pressure of 75 kPa and deviatoric stress of 10 kPa are also carried out. Through selected samples of different times in the process of creep to take microstructure test, the changed parameters of particles and pores are extracted to study the variation characteristics of microstructure in the process of creep. The results show that orientation of particles and pores is significant under the long-term loading conditions; the average pore diameter and average pore volume are gradually decreased; the specific surface area of structural undisturbed soil gradually increases; but the remolded soil is in the opposite trend; structural undisturbed soil particles and pore fractal dimension are decreasing, but the remolded soil shows the opposite trend.

The deep-sea sediment is exploited in the Pacific C-C district by our research group. Its physico-mechanical properties are studied by analysis of laser particle size, test of liquid and plastic limit, tests of nitrogen adsorption, consolidation and direct shear, and its mineral compositions and microstructures are analyzed by X-ray diffraction(XRD) and scanning electronic microscope(SEM) methods, respectively. Test results show that the deep-sea sediment is clay soil with predominant silt. It has properties of high liquid limit, high plasticity, high specific surface area, high consolidation coefficient, low compressive modulus and low shear strength, which easily results in slip and subsidence of the deep-sea mining machine. It is mainly composed of montmorillonite, quartz, feldspar, mica, chlorite and serpentine, with sheet-linked microstructure of silt and microbial debris filling with honeycomb flocculation microstructure of clay mineral together. The simulative soil substituting for the deep-sea sediment should be prepared based on the mineral composition, particle size, water content, specific surface area and shear strength as well as liquid limit, plastic limit and microstructure of the deep-sea sediment. The research results can provide an important basis for walking-characteristic study and optimization design of the deep-sea mining machine.

In the framework of classic elastoplasticity theory, it is assumed that rock material abides by the Mohr-Coulomb yield criterion and its plastic deformation obeys the non-associated flow rule during the process of strain softening. The strength parameters before and after the peak are obtained by analyzing the strain-softening behaviors of two sets of rocks in the triaxial tests. The piecewise linear approximation is made when numerical analysis is performed under the framework of the classical theory of plasticity. Then the strain-softening stage in the post-peak part of stress-strain curve can be simplified into a series of brittle-plasticity process. The stress-strain curve, lateral strain-axial strain curve, volumetric strain-axial strain curve and plastic volumetric strain-axial plastic strain curve are acquired by adopting the given strength parameter evolution laws and the strain-softening simulation method. The obtained stress-strain curves are in good agreement with those by the experiment, and the other curves are in the similar variation to those observed in experiments. The presented method and the results show that it is meaningful for determining strength parameters and studying the evolution law of the post-peak strength from laboratory test.

The geotechnical properties of the coral reef materials widely distributed in all islands in South China Sea have been extensively studied since 1980s in China. However, the majority of these researches are focused on the uncemented calcareous sandy soils and limited studies are available on the engineering properties of the naturally cemented calcareous soils such as calcarenites, which are the prevailing materials at the coral reefs in South China Sea. The state-of-the-art engineering practices in the calcareous soils worldwide indicate that the geotechnical properties of the cemented calcareous sandy soils are significantly different from those of uncemented coral reef soils. Considering the spatial heterogeneity of the naturally cemented calcareous soils, this paper presents factors that affect the strength properties from the microscopic studies. The correlations between the strength of the cemented calcareous sandy soil sample and various parameters, such as unit weight, cementation degree, void ratio, shape of grain as well as grain size, are established from the test results. The results show that good correlations exist between the strength with the unit weight and the degree of cementation of the calcareous sandy soils.

Four large-scale model tests are made to monitor the dynamic response of column-net structure subgrade. The influence of loading level, loading frequency and loading times on the dynamic parameters, the pile-soil load sharing and the strains of geogrid are studied. The results show that the dynamic stress and the dynamic acceleration increase fast in the bed of subgrade and underneath trend towards stability with depth after rainfall. During the stage of cyclic loading, the stress concentration ratio increases slowly. After rainfall, the stress concentration ratio can decrease about 56.4 % to 79.5 %. Compared with the construction stage, the strains of geogrid increase and the transverse distributions remain the same form after cyclic loading. The strains of geogrid among four piles within the width of subgrade increase fast after the formation of soil arching. The peak value points of strains of geogrid appear in the centre of the slope and subgrade, which emerges as W-shape in the cross-section. The strains of geogrid between two piles increase from the centre to the slope toe of subgrade, which emerges as stair-shape. The differences of strains of geogrid between the slope and the width of subgrade decrease gradually after the formation of soil arching.

The advantage of high quality assurance rate and good compression resistance of PHC pipe pile can be exploited by using them in the foundation of retaining wall of embankments. It puts forward a new research about the connection mode between PHC pile and retaining wall base. Based on in-situ tests in Guang-Qing Expressway extension project, the mechanism of the connection between the base of retaining wall and PHC pile is studied. Two connecting mechanisms, one with cover plate, called the cushion type and the other without cover plate, called embedded type, are studied. The research suggests that both cushion type and embedded connection between PHC pile foundation soil bear the main load, embedded links between pile and pile soil stress are larger, which are about twice of the cushion type, bearing capacity rate difference of inside and outside the pile for embedded type is more apparently. Deflection of the retaining wall cannot be prevented with the cushion type connecting mechanism, as opposite to the embedded type, in which the wall can be kept in a stable state by the constraint resulting from the piles. For both connecting types, the base of the retaining wall bears tension force along the embankment direction and non zero-compression zones exist in this surface. There is very little difference in horizontal displacement for two connecting types, and the horizontal load is mainly carried by the soils around the piles.

The conduits under high earth-fill have become more widely used in gully area in Western China; but the calculation of the circumjacent earth pressure on it is still not very clear. That needs further study. Based on the field tests on circumjacent earth pressure of the coal handling conduit in Shuozhou of Shanxi Province, the distribution and variation laws of the circumjacent earth pressures, stress state around the conduit are obtained. The test results show that there is a stress concentration for vertical earth pressure at the top of the coal conduit. The earth pressure increases faster when the filling height is lower; but the ratio of increase gets slower for the vertical earth pressure on the side of the conduit. The soil core method is preferably used when the filling height is lower; but the formula named ‘Gu Anquan’ is preferably applied to calculating the vertical earth pressure. Comparing culverts and underground penstock, the dimensions and rigidities of the high fill channels are larger, and using the coefficient of earth pressure at rest is recommended when calculating the coefficient of lateral earth pressure.

It is important to conduct the quality test on railway subgrade compaction in an efficient and reasonable way, which can ensure the project quality and improve the construction efficiency. Based on the test on the coefficient of subgrade reaction (K30) and modulus of deformation (Ev2 and Ev1), the characteristics of two small-sized plate loading tests are compared, the effect laws of test efficiency and test error through two loading modes of deformation stability and equal time intervals are mainly focused on, and the following conclusions are got Ev2, K30 can reflect the compaction degree of railway subgrade in a more direct way. The enduring period of each loading level has exerted great influence on the testing results and it takes longer testing periods to meet the 1% control standards of sedimentation stability. If the control standard of sedimentation stability of K30 test changes from 1% to 2%, or the interval of loading time sets at 6 minites, the testing period can be shortened about 30%-50% of the original time and the test error is within 10% of K30. The load-time control method is proposed in the subgrade reaction coefficient test, which fits the compaction quality test of railway subgrade made of gravel soil very well.

Based on many actual projects, through investigating slope rock mass containing multi-weak interlayers with different rock dip angles θ, slope angles β and structural face intervals h, safety factors and the location of critical failure surfaces are calculated by the method of Sarma and strength reduction method. Deformation failure mechanism of slope rock mass containing multi-weak interlayers is studied. The result shows: ①Failure mechanism and failure surface of layered rock slope are greatly different under the condition of different θ, it shows a progress of sliding failure, slip split failure, collapse, toppling failure, sliding failure when θ is increasing. ②As β and h are fixed, the stability of vertical layered slope is slightly larger than that of horizontal layered slope; and anti-dip layered rock slope is obviously larger than bedding rock slope. ③β affects the slope instability and damage features directly, and safety factor of bedding rock slope is reduced by nearly 53% as well as 40% reduction for the safe factor of anti-dip stratified rock. ④Effects of h are minor on slope deformation and failure mechanism but significant on stability, strengthening the monitor and reinforcing dense structure surface rock slope in practice are suggested.

In terms of surrounding rock control in large section open-off cut, open-off cut No. 5206 in a mine is selected as study case. The principal stress difference, deformation and fracture field are studied in the process of width from 6 m to 10 m with UDEC. Research results show that: the principal stress differences increase gradually in shallow part and decrease in depth. With the width increasing, the maximal principal stress difference maintains a constant level at first, then decreases gradually in the roof, and decreases at first, then remains constant in the floor, and increases gradually in the rib. The moving range of peak: rib>roof>floor. The curve of displacement in roof decrease exponentially, but has obvious inflection point, which decrease terracedly in the floor. The maximal surface displacement: roof>side walls>floor. The fracture field is divided into three areas: fracture transfixion area, fracture area and microfracture area, which are all distributed as semi-ellipse, the extent of cut through:middle part>two sides and shallow part>deep part. It is believed that high-performance bolt hold the slippage of fracture effectively. The double truss cable can be anchored in the shoulder angle area, which has no fracture, and pre-tension overlays the area. Also, the double truss cable can close the shallow part of roof and reduce principal stress difference. Based on above, a high strength and high pre-tension bolting support scheme of band net and double truss anchor are proposed. The open-off cut achieves self-stabilization after 10 days when it is dug out, the total roof separation is 3 mm, and the roof-to-floor relative convergence is 125 mm, both sides relative convergence is 94 mm.

Deep saline aquifer is a complex multi-components and multiphase flow system after CO2 injection. CO2 advection happens because of pressure gradient and velocity difference and dries out water in saline aquifer, which changes phases of the components and causes salt precipitation. The precipitated salt clogs the pore space in aquifer and decreases CO2 injectivity, so it is very important to research the effect on dry-out problem and site selection. 2D radial model is built and multiphase fluid flow function is established. By considering relative permeability function and capillary pressure function, the effects of CO2 injection rate, salty in aquifer, input maximum capillary pressure and characteristic parameters of capillary pressure on dry-out problem are studied. The dry-out problem can be quantitatively described by solid saturation. Results show that CO2 migration separates the aquifer into three domains: dry-out domain, gas and liquid mixture domain and liquid saline aquifer domain. The dry-out problem happens nearby the injection well. If capillary pressure is considered, the solid saturation will increase when injection rate decreases, salty in aquifer increases, or capillary pressure increases. As a result, increasing CO2 injection rate, injecting water into aquifer before CO2 injection, and decreasing capillary pressure by choosing a large pore size homogenous formation can avoid the clogging of pore space with salt precipitation, which enhances CO2 injectivity.

At present, there is a large-scale construction of salt rock storage groups in China. Many optimizable factors exist in the planning and design phase with different influencing magnitudes. The three main factors, i.e. the geometric distribution of ellipsoidal caves, the width of pillars and the position of the interlayer, are selected to be studied. The purpose of the research is to analyze the influence degree of the three factors and to determine the optimal storage group form. Three levels are set for each factor and nine text schemes are made based on the principle of orthogonal experimental design; and the related numerical simulation experiments are conducted. The improved experimental index of displacement and stress is chosen to evaluate laminated salt rock storage group. Using extreme difference analysis and variance analysis, the test results are analyzed in detail. The research results show: among the three factors, the position of the interlayer has the most significant influence, the width of pillars is in the second place and the geometric distribution of ellipsoidal caves is the third. The optimal storage group form: the angle of the adjacent caves is 60°, the width of pillars is 2.0D and the interlayer is distributed in upper 1/4H of the cavity.

Now in China the tunnels excavated by TBM become deeper and longer than before; and TBM jamming accidents are very common. In order to reduce these accidents, comprehensive evaluation of TBM risk accident induced by tunnel deformation can be carried out in advance. Therefore, the research on comprehensive risk is started. First, the calculated model of comprehensive risk probability is established by probability theory. According to the derived model, the accident probabilities of different consequence ranks can be calculated. Second, the ratio of jamming time and pure penetrating time is used as graded index. The index combined with the data gathered from previous TBM construction is used to divide consequence ranks of construction time delay. Then the comprehensive risk evaluation criterion for deformation is established. Finally, the jamming risk in certain parts of practical project is analyzed by the theory proposed above in two different construction conditions. It is demonstrated from the calculated conclusion that the construction risk is bigger when no measures are taken; however, the risk can be controlled within the acceptable extent when proper measures are taken.

Conventional soil stabilization methods usually involve the addition of some agents or chemicals into soils to generate a composite soil mass with high strength. However, the application of these soil stabilization methods, in general, requires a very long curing time; and it is particularly not economical. By focusing on three typical local soils in Hong Kong, namely, marine clay, completely decomposed granite (CDG), and completely decomposed tuff (CDT), an experimental study of soil improvement is presented, with the use of an innovative and practical bio-enzyme-based soil stabilization technology. As suggested by the experimental results, the stabilization effect of bio-enzyme on marine clay is the best, and unconfined compressive strength of marine clay is able to increase by as high as 20%; while the improvement effect on CDG soils seems not to be quite good since shear strength indices have merely slightly changed after bio-enzyme treatment. It is to be especially noted that the compressibility of marine clay and CDT is improved after treatment, instead of being reduced. Furthermore, the fundamental mechanisms of the bio-enzyme-based soil stabilization are clarified from a microscopic perspective. It is pointed out that the stabilization effect mainly results from the bonding effect generated by the interaction between clay mineral molecules in soils and bio-enzyme molecules.

A principle of nonlinear relationship proposed by Hoek is elaborated, then based on that, a new algorithm of obtaining shear strength of rock mass is brought forward by considering the overall or average level when rock mass is broken. Through the new algorithm, the magnitudes of shear stress can be calculated at different given normal stresses, and a series of shear stress and normal stress are generated; then the values of cohesion and internal frictional angle are determined via linear regression analysis, which solves the determination of shear strength under certain normal stress that Hoek-Brown criterion fitting algorithm can not do. Then a research on shear strength of rock mass from China-Myanmar oil and gas pipelines (domestic section) Lancang River across domain engineering as an example and a further comparison study is taken. It is shown that the values of cohesion and internal frictional angle determined by the first method according to the new algorithm are little different to these gained through the Hoek-Brown criterion fitting algorithm. The reason is that the factors considered are not the same, while the normal stress is in considered by the two algorithms resulting in confining pressure and normal stress are varied practical engineering is low, the cohesion and internal friction angle proposed by the second method according to the new algorithm, the results obtained are close to the actural situation, which has overcome the problem that Hoek-Brown criterion fitting algorithm relies on the confining pressure.

Gully bed erosion depth is one of the most important parameters for design of debris flow control works. So far, there are few researches on gully bed erosion depth and there are no good methods for calculating the gully bed erosion depth. How to calculate the gully bed erosion depth is also a technical problem which urgently needs to be solved for debris flow preventing engineering design. The driving forces of viscous debris flow and potential resistance which may take part in the motion are analyzed. And then the calculation formula of the maximum erosion depth is obtained. The formula shows that the erosion depth of gully bed increases with the increasing of depth and density of debris flow, longitudinal slope of gully bed and viscosity of gully bed materials, while decreases with the increasing of internal frictional angle of gully bed materials. Comparing with the existing calculation methods, the formula is based on a strict theoretical derivation, the calculation result is more accurate and can be used to calculate the erosion depth in the occurrence of debris flow area for different frequencies, and provides a technical support for the design of viscous debris flow control works. The practical value of the calculation formula is illustrated.

Complex foundation pit has explicit three dimensional effect, and the three dimensional stability of the pit cannot be obtained exactly with present computational methods. The space failure range also cannot be defined easily. The Bezier bicubic curved surface is adopted to fit the displacement isosurface which passes the nodes; and it is assumed that the displacement isosurface is a local failure surface of the pit slope. Then the normal stress vertical to failure surface and the shear stress along the slip direction are calculated. A new point safety factor is defined as the ratio of shear strength at the point and shear stress along the slip direction; and the whole safety factor is the weighted average value of the point safety factor according to the area. The stability of pit slope in a normal university is analyzed. The results show that point safety factor has the characteristics of small values in the middle and higher values at two ends along the profile. The region with the values of point safety factor less than 1.06 extends from the bottom to the top of the slope, which indicates the location of slip surface. The dangerous slip surface is the displacement isosurface with values of 15 mm. The consistency of computational results along the specified profile by the methods of rigid equilibrium method and point safety coefficient has also verified.

The stress-strain-time relationship of the Koppejan creep model based on consolidation tests is improved to the incremental form, which is generalized to 3D stress space, and then the improved Koppejan creep model for soft soil is established in this work by the incremental form. Based on the ABAQUS software, an UMAT is developed by the improved Koppejan model. Using long-term consolidation tests of Shanghai soft clay to validate the model, the results show that the Koppejan model describes the creep property of Shanghai soft clay very well; the computational accuracy of the improved Koppejan model is better than that of the classical Koppejan model, and the two models work better for mucky clay than silty clay. Using the improved Koppejan creep constitutive model to the three-dimensional FEM of a high-speed railway bridge pile foundation long-term deformation, and comparing with the measured settlement, the results indicate that the improved Koppejan can be well used for computing long-term deformation of actual projects.

The time-domain explicit finite element method for the wave motion of saturated two-phase medium is conditionally stable, and the stability of this kind of method can be evaluated by the transfer factor of transfer matrix of calculating formula. The smaller the value of transfer factor is, the better the stability of the explicit finite element method is. The stability of two time-domain explicit finite element methods for the wave motion of saturated two-phase medium is compared based on the transfer matrix of the calculating formula of the method, as various values of physical parameters of saturated two-phase medium are adopted. Calculating results show that the value of permeability coefficient has a remarkable effect on the distinction between the stability characteristic of the two methods. When a large value of permeability coefficient is adopted, the stability characteristic of the two methods does not distinguish obviously from each other. On the other hand, when a small value of permeability coefficient is adopted, the stability characteristic of the two methods will have a remarkable difference from each other. One method has more ascendant stability characteristic than that of the other method.

Three-dimensional finite element model of breakwater-ground system is established for calculating the influence of traveling wave excitation based on elastoplastic constitutive model. Considering the dynamic breakwater-soil interaction, the model is solved by supercomputer based on contact balanced algorithm for explicit parallel computing. The difficulty of calculating the large-scale nonlinear finite element problems is solved based on this method. The settlement of wave barrier, the permanent deformation of breakwater and the bending moment of wave barrier under different wave velocities are researched. The results show that when considering the wave travelling effect of ground motion，the settlement of wave barrier increases with the wave velocity increasing, and gradually approaches the response of uniform excitation. The travelling wave effect has little effect on the residual deformation mode of breakwater, and the deformation mode under different wave velocities are almost the same. The permanent displacement of breakwater under SH wave excitation is bigger than the one under P wave excitation. In most cases, the bending moments of wave barrier under nonuniform excitation are bigger than ones under uniform excitation. The results are valuable for the choice of aseismic measures for breakwater.

Upper bound solution of limit analysis with rigid blocks is widely applied to geotechnical engineering stability analysis. However, rigid blocks failure mechanism and complicated geometric relation are needed to assume and analyze for different geotechnical subjects. In order to banish these restrictions, a finite element upper bound solution with rigid translatory moving elements based on nonlinear programming is presented. And the key problem of determine initial value for nonlinear programming model is also solved. This method firstly discretizes computation domain into rigid elements, then defines the element velocities and nodes coordinate as decision variables, and uses upper bound theory to establish nonlinear programming model to acquire a solution. Corresponding finite element upper bound program is compiled to verify stabilities of slope and shallow tunnel to illustrate that velocity discontinuities of elements adjust automatically into best directions to acquire failure mechanism with distinct characteristics and upper bound solution with high accuracy. The method is expected to be applied widely to stability analysis for slopes, tunnels and other geotechnical subjects.

Discontinuous deformation analysis (DDA) is a newly developed discontinuum numerical method for simulating large deformation and displacement of block systems. The core research of this thesis concentrates on improving the method to enforce contact constraints in DDA. The penalty function method and Lagrange multiplier method or its variants are generally utilized to enforce contact constraints in DDA. But it is difficult to set a reasonable penalty parameter. Furthermore, discontinuous change in stiffness matrix due to the open-close iteration frequently causes poor convergence. To avoid troubles in setting artificial penalty parameters and open-close iteration, we reformulate discontinuous deformation analysis method as mixed linear complementarity model (LCDDA). On the basis of these, both advantages of local quadratic convergence rate of the Newton method based on nonsmooth analysis and global linear convergence rate of steepest descent method are integrated to set up the algorithm. According to the above achievement, the complete code for LCDDA is accomplished. Some examples are analyzed to verify the precision and feasibility.

In a one-dimensional consolidation equation, the permeability coefficient is the function of matric suction in an unsaturated soil. The consolidation equation in one dimension is strongly nonlinear due to the presence of the permeability coefficient. As an analytic solution to nonlinear problems, a homolopy analysis method (HAM) is efficient in the selection of series basis functions and auxiliary linear operators, and easily make the solution convergence. In order to analytically solve the equation, the HAM was introduced in the present study. During the solution, the pore air pressure was assumed as the atmospheric pressure. The equations with two unknown variables were then reduced to dimensionless pore water pressure as only one basic unknown variable subjected to constant pore air pressure. Firstly, a governing equation in a dimensionless form was derived from the basic one-dimensional consolidation theory after the integration with the Richard empirical formula. The method was then used for a mapping technique to transfer the original nonlinear differential equations to a number of linear differential equations. These differential equations are not dependent on any small parameters, which is convenient to control the convergence region. After this transferring, a series solution to the equations was then obtained by the HAM after selection of auxiliary linear operator parameters. Finally, comparisons were carried out between the analytical solutions and the finite difference method in case of compacted kaolin. It can be found that the series solutions indicate that the pore water pressure increases firstly, and then decreases with the depth after the consolidation of the compacted kaolin. The results indicate that the analytical solution in the present study is reasonable.

Three-dimensional bounding surface constitutive model for rockfill materials incorporating the general nonlinear strength criterion can well capture such behaviors of rockfill materials as the strain hardening, the post-peak strain softening, and the volumetric strain contraction and expansion in the three-dimensional stress space, which cannot be predicted by the traditional constitutive model proposed before. Based on the theory of 3D bounding surface constitutive model and according to the further developing platform provided by FLAC3D, the 3D bounding surface model is successfully developed in FLAC3D with VC++ environment; and then key steps, programming essentials and methods of constitutive model’s secondary development are given based on FLAC3D code. The calculative results of 3D bounding surface model using FLAC3D under general triaxial compression tests are compared with the theoretical values. The results are turned out that the calculative results can aptly fit the theoretical values and predict the behaviors of rockfill materials. In particular, model parameters of bounding surface model are simple and can be determined by general triaxial tests. Therefore, we can see that it shows the superiority and reasonableness of 3D bounding surface model’s procedure.

In rock-type materials, two types of cracks are observed under combined compression and shear loading: the tensile wing cracks, and the frictional secondary cracks. On the basis of the extended finite element method (XFEM), an extended finite element displacement enrichment method to model the interaction of the frictional crack surfaces is given. Besides, an approach to deal with the branched cracks is also presented. The maximum circumferential tensile stress criterion and the Mohr-Coulomb criterion are adopted respectively to predict the initiation and propagation of the tensile cracks and the frictional cracks. A Matlab program Betaxfem 2D has been developed and the proposed methods are verified through two examples; and then the crack bifurcation and propagation processes of the specimens with pre-existing closed fractures under uniaxial compression loading are simulated. The comparison with the experimental results indicates that methods proposed can be used to simulate and predict the initiation and propagation of the interactive tensile and frictional branched cracks in rock-type materials.

Based on the explicit finite element method and 32 CPU parallel computing cluster platform of ABAQUS, considering the heterogeneity of ground soils, a large-scale refined two-dimensional finite element nonlinear analytical model for a multi-geomorphic composite site that consists of alluvial flat，terraces，undulating hilly terrain and monadnock is established in the time domain. The SH wave is taken as the bedrock earthquake motion and nonlinear seismic effect characteristics of the composite site are analyzed. The results show that: (1) the peak ground acceleration(PGA) and peak ground displacement(PGD) of different observation points on the ground surface vary with each other, which is caused by the terrain differences. When it’s under the same input ground motion, the ground surface PGA of alluvial flat area shows obviously larger than that of the first terrace area, however, the PGD of alluvial flat is smaller than that of the first terrace, and the PGD difference is relatively smaller in the alluvial flat area. (2) The horizontal acceleration response spectrum of ground surface appears a double-peak or multi-peak phenomenon, the amplification of ground motion seems outstanding in the frequency band from 0.4 Hz to 1.2 Hz. Moreover, the acceleration response spectrum is related with the input ground motion. (3) The amplification effect and focusing effect of seismic motion are more obvious in specific frequency band(0.5 Hz to 1.75 Hz), when the frequency is lower than 0.2 Hz or higher than 2 Hz, the seismic amplification characteristics are not apparent. (4) The location of observation point has influence on the acceleration duration，the changing terrain may cause difference of seismic motion duration，besides，the motion duration, in some way, is dependent on the input ground motion characteristics.