When a gravity structure, such as a caisson or a jackup spudcan, is pulled out from the saturated soft clay foundation, a negative pore pressure will be developing, which may greatly increase the pullout force. Therefore, it is very important to estimate the negative pore pressure developed in the soil during the unloading process. The development of excess pore pressure is studied based on the triaxial extension tests under the isotropic consolidation conditions. By combining the modified Cambridge model with the Henkel’s approach, the mechanism of negative pore pressure generation is revealed and a method is developed to predict the magnitude of the negative pore pressure. Meanwhile, the development rule of pore pressure under the isotropic consolidation conditions is obtained and explained why there will be a change of pore pressure from negative to positive. The research indicates that under the unloading conditions, the generalized shear stress induces positive pore pressure and the averaged principal stress generates negative pore pressure in the soil samples. The superposition of these two effects results in negative pore pressure in the initial unloading stage, and then becomes positive with the increase of strain. The research results can be used to predict the required force to pull out a structure from saturated clay foundation.

In order to study the dynamic responses of anchorage system under earthquake，the model of landslide supported by anchor bars and lattice beam is carried out using shaking table. The difference of dynamic response of anchor bars in different locations and the mechanism behavior of anchor bars under earthquake are studied by inputting waves of Wenchuan earthquake, EL Centro earthquake and sine seismic wave to monitor the axial force of anchor bars, landslide acceleration and landslide displacement. The results suggest that under low level earthquake loading (0.05 g～0.40 g), the surface effect of slope toe is more obvious than other parts of the landslide, and the sliding surface near slope toe is prone to fracturing because of repeated rubbing and cutting. Based on pseudo-static method design, it is advised that short anchors or soil nails should be driven into toe slope in order to balance the uncoordinated reciprocating motion of sliding surface near slope. Under high level earthquake loading (0.6 g～0.8 g), the surface effect of slope shoulder is more obvious, the fracture in sliding surface near slope top is longer and wider than before, and the sliding surface is more likely to crack or collapse. It is advised that the first-row anchors should be extended in order to restrain the development of cracks in slope sliding surface. Meanwhile, using short anchors or growing plants with developed root system at top of the slope is necessary to protect the slope. Slope with anchorage system is not prone to as-a-whole slope damage like plain soil slope, and under high level earthquake loading, firstly anchors in slope toe are damaged, and then tensile force of top anchors sharply rises due to cracks caused by pulling force at the top surface of slope. The results provide a reasonable foundation for the design of anchor bars under earthquake.

This paper aims to study the softening characteristics of gypsum surrounding rock in the construction process of the tunnel. The multi-factor orthogonal softening experiments were carried out on gypsum rock from Lower Triassic Jialing River Formation. Then we analysed the effects of immersion time, solution temperature and concentration on mechanical parameters of gypsum rock after softening, respectively. The results showed that both the softening coefficient and the elastic modulus decreased with the increase of immersion time, furthermore their decreasing rates showed the same patterns. It was also found that they linearly decreased with increasing the solution temperature. Among these three influencing factors, the immersion time played the most critical role in softening experiments, and its weight was up to 80%. However, the effect of temperature ranked second with its weight of about 16%, and the effect of concentration was very slight. The prediction formulas were established for the elastic modulus and the uniaxial compressive strength of gypsum rock under the coupling action of the immersion time and the solution temperature. By analysing the experimental results, the construction methods and suggestions for gypsum rock tunnels were put forward from the aspects of the division of surrounding rock hardness, prevention and drainage measures and the supporting measures.

With the growing demand of clean energy, much attention has been paid to the supercritical CO2 strengthened shale gas development technology. This paper aims to investigate the effect of supercritical CO2 on the mechanical properties of the shale. Supercritical CO2 immersion experiments were carried out on the Silurian shale from Sichuan Basin Longmaxi for different durations. Besides, indirect tension tests (i.e., Brazilian disc tests) and triaxial compression tests were conducted to obtain the variation regularities of the strength and deformation of shale before and after immersion. Combing with the results of XRD, porosity and SEM tests, the effect of supercritical CO2 on the mechanical properties of shale was preliminarily discussed. The results showed that after supercritical CO2 immersion, the tensile strength, triaxial compressive strength, and elastic modulus (Young’s Modulus) of the shale all decreased at various degrees. However, the amount of shale strength loss increased with increasing the immersion time. Meanwhile, XRD, porosity tests and SEM results indicated that supercritical CO2 degraded mechanical properties of shale by changing the structure of mineral constituent, reducing the degree of cementation, and changing the particle framework and pore structure.

Rubber-sand mixture (RSM) has long been recognized as a light energy absorbing material with wide applications in engineering. The moisture content of RSM may vary depending on the application and external conditions. Therefore, it is important to investigate the influence of water content on dynamic shear modulus and damping ratio of RSM for its appropriate application. Shear tests using large-scale cyclic SGI simple shear test apparatus were carried out on RSM with 6 kinds of water content under 3 kinds of vertical pressures. Test results show that RSM dynamic shear modulus and damping ratio vary with water content when the cyclic shear strain amplitude ranges from 1% to 3%. Test results indicate that: The decrease of dynamic shear modulus of pure sand with increase of water content is restrained by adding rubber particles. When the rubber content reaches 30%, the dynamic shear modulus of RSM shows an increasing trend with the increase of water content, which is contrary to the case of pure sand. With the increase of the rubber content, the damping ratio of RSM shows a decreasing trend with the increase of water content. Though dispersion of results is present, the effect of water content on the dynamic characteristics of RSM is significant. Specifically, the average influence of water content on dynamic shear modulus of RSM reaches up to 20%, and the maximum average variation of damping ratio of RSM shows an absolute value 0.025 under the influence of water content. For the stability of dynamic characteristics with water content variation, a rubber mass content of 20% in RSM may be the optimum content.

Tailings dam is in danger of liquefied instability under earthquake. The liquefaction process can be indirectly reflected by the evolution rule of the dynamic pore-water pressure. To study the dynamic pore-water evolution, a series of dynamic triaxial tests is conducted. The results show that: the development of dynamic pore-water pressure of the tailings is characterized with different stages, including an S-shaped curve under isotropic consolidation and a J-shaped curve under anisotropic consolidation. The critical pore-water pressure of tailings is close to confining pressure under isotropic consolidation. The critical pressure, less than confining pressure under anisotropic consolidation, decreases with the increase of confining pressure and consolidation stress ratio, and increases as the average particle size of tailings increases. The critical pore-water pressure equations under the condition of isotropic consolidation and anisotropic consolidation are derived based on the theory of limit equilibrium, which explain the dynamic evolution of pore-water pressure observed in dynamic triaxial tests. The results provide reference for seismic design of upstream tailings dam in earthquake area.

Gudehus-Bauer hypoplastic constitutive model and the determination of its parameters are studied in the paper. Confined compression test is used to determine the parameters hs and n of hypoplastic constitutive model for rockfill materials. According to the derivation of the model, the relationships between the parameters , and the confining pressure are established, and new fitting parameters are proposed. Considering the dilatancy for rockfill materials, a new term of volumetric strain is added into the linear term in the model with a good fit to the axial strain-volumetric stain curve of triaxial consolidated drained tests. The new parameters of the model and the improvement on Gudehus-Bauer hypoplastic constitutive model are verified by means of rockfill large-scale confined compression test and triaxial consolidated drained tests. Compared with the test results of rockfill materials, the present model and its parameters can simulate the stress-strain characteristics of rockfill materials, and can improve the simulated axial strain-volumetric stain curve. Finally, the simulation to the triaxial consolidated drained test in loading-unloading condition by the improved Gudehus-Bauer hypoplastic constitutive model is carried out, and the results show that the improved model has a certain application of unloading.

It’s important to study the variation of shear strength and parameters of pile-soil interface for clayey soil with the excess pore water pressure in engineering practice. A series of direct shear test with different interface roughness, water contents and shearing rates is performed using the large-scale direct shear apparatus with constant normal stiffness. The effect on excess pore water pressure is examined under different test conditions. Influence of excess pore water pressure on shear strength of pile-soil interface in clayey soil is obtained. The results show that the excess pore water pressure decreases, and the shear strengths of pile-soil interface, effective cohesion and effective coefficients of friction increase with the increasing interface roughness. The excess pore water pressure increases and the shear strength of pile-soil interface decreases with the increasing water content. The main influence of increasing water content on shear strength of pile-soil interface is changing the cohesion. The cohesion first increases and then decreases. The influence on the coefficients of friction of pile-soil interface is little. The excess pore water pressure increases and the shear strengths of pile-soil interface also decrease with the increasing shear rate. The cohesion of pile-soil interface first increases and then decreases, but the variation is less than 2 kPa, and the little influence on the coefficients of friction of pile-soil interface. The change of shear strength and parameters is influenced by variation of excess pore water pressure, depending on interface roughness, water contents and shear rates. This experimental research is useful for relative engineering design.

During earthquakes, the shear wave can be simulated by the horizontal bidirectional shear stresses mutually coupled in magnitude and direction. To investigate the influence of cyclic shear stress ratio (SSR) and phase difference on dynamic characteristics of soft clay under bidirectional loading, 19 groups of undrained cyclic simple shear tests have been performed on typical Wenzhou soft clay by using multi-direction cyclic shear apparatus. The test results show that the threshold cyclic stress ratio SSRx,menkan for the typical Wenzhou soft clay under unidirection cyclic shear loading is 0.096 and the critical cyclic stress ratio SSRx,limit is 0.15. Under the bidirectional cyclic shear loading, the increase of cyclic number N and the improvement of SSR will promote the development of shear strain and pore pressure u. When SSR is slightly larger than SSRx,menkan, the increase of will decrease the development of and u. When SSR is between SSRx,menkan and SSRx,limit, shear strain and pore pressure u increase rapidly with the increase of . When SSR is larger than SSRx,limit, the effect of phase difference on the shear strain and pore pressure u can be ignored.

By using the large physical model test, this paper investigated on the mechanism of the component steel tube anti-sliding piles distributed in different spacings. The authors have compared factors affecting of the spacing on the mechanism of the anti-sliding among the pile top’s distance, the cut-export distance, the bending moment distribution under changing loads and the destruction situation. The experiment results show that when the spacing is increased from 18.7 cm (3.9 times of d the pile external diameter) to 23.5 cm (4.9d) and 28 cm (5.8d), the distance of pile top and cut-export increases gradually. Besides, the bending moment of the pile also increases gradually. But when the spacing is increased from 3.9d to 4.9d, the increase range of the distance and the bending moment are greater than that when the spacing is increased from 4.9d to 5.8d. Thus, during the design of the component steel tube anti-sliding pile, it is better to increase the spacing to 5.8d or more when the landslide thrust is small and decreases the spacing to 3.9d or less when the landslide thrust is large. Whether the landslide thrust is large or small, the component steel tube anti-sliding pile has a low-cost performance to treat landslide when the spacing is 4.9d.

Based on the static and dynamic centrifugal model test of cantilever anti-slide pile in accumulation landslide, the mechanical characteristics of anti-slide pile and the seismic responses of landslide were analyzed using data from soil-pressure sensors, strain gauge pairs of pile and accelerometers. The results show that the distribution laws of earth pressure on the back of piles and bending moment of pile under static loading and dynamic conditions were different. The static earth pressure of pile back was significantly greater than that induced by earthquake action, but the static bending moment of pile was far smaller than that caused by seismic vibration. The dynamic earth pressure on the back of piles and the bending moment of piles increased with the increase of peak ground motion. The action points of the maximum earth pressure on the back of piles and bending moment of pile were lower under seismic actions than under static condition. There is superficial amplification effect near the crest of landslide and elevation magnifying effect inside the landslide. The phenomenon of wave type transform was significant, to which more attention should be paid in aseismatic design.

The standard of deformation stability and the design of stepwise loading are the key factors affecting the efficiency and accuracy of K30 test, and closely related to deformation time-effect state of the filling in the loading process. To get time history curve and deformation data, three small plate loading tests of unit structure filling models are carried out in optimal water rate at compaction coefficient (K) of 0.90, 0.95 and 1.00. The change of the deformation state of the soil during the loading process of K30 test and the influence of the loading stability time are discussed. The results indicate that, the deformation of well compacted silty clay filler in the process of K30 test is mainly elastic, and almost in state of slow convergence. With the increasing of load, the ratio of plastic deformation gradually develops, and the evolution trend of deformation state is presented from weak to strong. When accumulated to 1.25 mm, the deformation is in the sub-state of weak in slow convergence. The standard of “0.01 mm/min”, step-loading time of which shows accelerating growth pattern, can apparently reduce test time with acceptable test error. On the principle that the deformation state should be in apparent state during loading process, and deformation should exceed 1.25 mm, for loading system of five steps and 0.04 MPa load increment, it is suggested that the K30 value in the detection of compaction quality of fine grained soil filler should be between 60-160 MPa/m.

A piled embankment with reinforcement is one of embankments widely used by high-speed railways. The effect of vibration velocity response under train loads of high-speed railways on train running safety and surroundings is crucially important. A large-scale piled embankment model was set up with X-section cast-in-place concrete pile as vertical reinforcement. The vibration performance of the piled embankment under train load of high-speed railways was investigated. The influence of loading frequencies, loading amplitudes, and the number of cycles on vibration velocity of the piled embankment model was revealed in this paper. The results show that it is faster for attenuation of vibration velocity along the transverse direction of the piled embankment. The impact of the embankment on energy dissipation is remarkable, so the effect of vibration induced by train running on surroundings is weaker. The attenuation of vibration velocity along the piled embankment depth is different from the transverse direction of the piled embankment. The piled embankment system shows the greatest attenuation of vibration velocity is from track slab and reinforced concrete base to surface layer of subgrade bed. The gravel cushion with geogrids attenuates less due to shock isolation cushion. The piles in the foundation make attenuation of vibration velocity by train running slower and influence depth deeper. Vibration velocity amplitude increases gradually with the increase of loading frequency, and reaches a peak at 25 Hz. It becomes bigger step by step with the increase of loading amplitude and stays the same with the increase of the numbers of cycles.

In order to explore salt expansion properties of sulfate saline soils under the decrease of water content, 11 sets of indoor simulation tests under one time decrease of water content, lasting for 2 868 hours, have been carried out using the self-developed experimental equipment. It is shown that the salt content of swelling initiation of sulfate saline soil is 1.2%. The volume of sulfate saline soil shrinks when the salt content of sulfate saline soil is less than 1.2%, and there is an approximate second-degree lower convex parabola correlation between salt expansion rate and salt content. The drying shrinkage rate of sulfate saline soil reaches the maximum value when the salt content is 0.5%. When the salt content is greater than or equal to 1.2%, the volume of sulfate saline soil swells. The salt expansion rate increases with the increase in salt content and there is a linear correlation between the two. The sensitive water content range of salt expansion, slightly influenced by salt content, is generally stable when salt content increases from 1.5% to 4.0%. It can also be concluded that the decrease rate of water content is negatively correlated with the salt content, and the decrease rate of water content in the same salt content decreases with the decrease of water content. Based on the above test results, 4 sets of salt expansion tests have been carried out under single cooling. It is shown that the temperature range initiating swelling of sulfate saline soil gradually goes up and the interval range of salt sensitive temperature range gradually expands with the increase in salt content. According to the test results from two kinds of test conditions, the salt expansion rate of single cooling is higher than the salt expansion rate of one time decrease of water content. However, the difference between the salt expansion rate of the two testing conditions is basically constant with the increase in salt content. It is expected to provide guidance for evaluation on salt expansion of the foundations of residual sulfate saline soil with high salt content in the northwest region of China.

Defected rock mass seriously threatens the safety construction and long-term operation of underground engineering. It is significant to investigate the physical variables before and after the instability, which provides useful information to identify the precursor of rock mass instability failure. This study is to conduct experiments on the gypsum in combination with crack elements, and further to investigate its space-time evolution process of the failure-instability. Before the instability failure, the changes in the corresponding physical field are captured and compared at critical moments. The processes from the independent activity to overall synergistic movement of rock mass are discussed as well. Experimental results indicate that the complete destruction of the crack combination can be divided into three stages: strong deviation from the linear, static sub-instability and dynamic sub-instability. In the non-linear stage, the relatively strong segment become the main bearing body. The isolated strain accumulating area and releasing points appeared firstly in the weak segment, and then acted on the interface between strong and weak segments. Strain accumulating areas in the weak segment rapidly increase, extend and migrate. Different lithologies of rock mass present strain synergistic transmission with unitary and regional characteristics. Finally, strain releasing points are mutually connected. When both the synergistic effect and strain accumulation degree in the strong segment reach their limits, the stage of rapid destruction of rock mass into full instability emerges. A stope fault model is established to further verify that the results are scientific and rational. This model was based on the engineering background of water inrush in a coal mining. The two-axial loading experiment was carried out to investigate the process of water inrush and trigger conditions caused by the fault instability. The mechanism of synergetic failure in the sub-instability stage is revealed. Experimental results indicate that the sub-instability stage of jointed rock mass makes the beginning of synergistic failure.

This study is to investigate shear properties and rock block breakage characteristics of soil-rock mixtures (SRMs) by lab large-scale direct shear tests and particle discrete element numerical simulation using PFC2D. A particle discrete element numerical method for SRM, accurately describing the shape characteristics and the breakage effect of rock block, is presented with comparisons of the results of large-scale direct shear tests and sieving analysis tests. The shear properties and rock block breakage characteristics of SRMs of six different rock block contents under four different normal stresses were studied. The result basically complies with Mohr-Coulomb strength criterion, showing the higher the rock block content, the higher the shear strength of SRMs. With the increase of rock block content, the internal friction angle of SRMs shows slow-fast-slow growth trend, but the cohesion presents a trend of first increase, then decrease and then increase. The breakage forms of rock block can be classified into 4 categories: surface grinding, partial breakage, complete rupture and complete breakage. A new particle breakage index is presented to accurately reflect the breakage degree of rock block with particle size greater than 5 mm. The index increases as the rock block content and normal stress increasing. The analysis of shear surface of particle discrete element numerical model shows that the shear surface appears ‘shear oblique’ phenomenon. The breakage degree is more obvious and the number of shear cracks near shear surface increase as rock block content increasing. The failure of S-RMs during shearing is tension-shear mixing failure.

Distributed optical fiber sensing technology was applied to the quantitative detection of intercalated mud bored pile. One dimensional model was proposed to simulate the heat transfer of single-layer cylindrical wall made up of heating fiber and pile body medium . Mud content of pile was detected by the variety of temperature rise of optical fiber in the pile. Intercalated mud pile model tests of optic fiber detection were designed with the mud contents of 0, 33.3%, 50%, 66.7% and 100%. After the calibrating of fiber optic sensor, the temperature rise and thermal conductivity coefficient of piles were investigated under appropriate range of heating power and heating time. It was founded that the rate of temperature rise of the fiber was increased with the mud content under the same heating power. Further analysis showed that there is a good linear relationship between mud content and thermal conductivity coefficient of pile. At last, the accuracy of this technique was analyzed by acoustic wave test. The results showed that the accuracy of optic fiber detection method is close to that of the acoustic wave detection method when the mud content of pile is more than 10%. This studies provided a way for quantitative detection of pile integrity.

Rock is characterised as the elastoplastic coupling, which means its elastic parameters change with the variation of plastic deformation beyond the threshold of yield. Based on current classification, definition and constitutive function of strain, the following work was conducted under the framework of elastoplastic coupling. Firstly, through compressive tests, the evolution of elastic parameters was analysed from the aspects of the damage and plastic deformation. The strain was mainly studied under the condition of loading incremental step. It was suggested that the strain increment corresponding to the loading increment could be decomposed into elastic strain increment , elastic strain increment and plastic strain increment respectively. Furthermore, was following the generalised Hooke’s law, while was against Hooke’s law. The elastic relationship was constituted with elastic parameters after loading incremental step. Constitutive functions were then formulated in strain and stress spaces under loading, neutral loading and unloading conditions, respectively. The discussion was made on the internal variable increment reflecting the microstructure change. In general, the classification and definition of strain are explicit in concept and suitable for strain hardening and softening sections. Moreover, the elastoplastic coupling constitutive function considers the influences of different loading conditions on the changing rates of elastic parameters. Overall, the strain definition and constitutive function are in accordance with rock elastoplastic coupling characters.

A series of laboratory tests is performed to study the influence of drainage condition and coarse grain content on the dynamic residual deformation characteristics of the saturated gravel soil. Extensive tests are carried out under no-drainage, single-drainage and double-drainage conditions respectively, by using the improved GDS cyclic triaxial testing system. The gravel soil samples consist of five different gradations. The results indicate that, the dynamic residual shear strain is dominated by both the coarse grain content (the content of coarse particles, whose diameters are bigger than 5 mm) and the drainage condition. The residual shear strain significantly decreases as the coarse grain content and the number of drainage surfaces increase. When the number of the cyclic loading reaches 30, the dynamic residual shear strain under the no-drainage condition is 2-3 times that under the single-drainage condition, and 4-9 times that under the double-drainage condition. However, the dynamic residual volumetric strain of the saturated gravel soil decreases with the increase of the coarse grain content, as well as the number of drainage surfaces. The corresponding dynamic residual volumetric strain under the double-drainage condition is 2-2.5 times that under the single-drainage condition.

To study macro and micro mechanical properties of cone penetration tests in structured sand, a pure sand ground was prepared firstly under 10 g condition. Then, a bond contact model considering bond thickness was implemented into the pure sand ground to prepare the structured sand ground. A penetrometer composed of four sides of rigid walls was moving downward to simulate cone penetration tests in structured sand ground. The results showed that with the increase of penetration depth, the tip resistance increased gradually but the growth rate slowed down step by step. When the tip reached the critical depth, the tip resistance began fluctuating around a certain value. The contact force chain around the tip is very concentrated. With the increasing penetration depth, the concentrative degree of contact force chain increased gradually and the concentration range of contact force chain expanded gradually. In the process of penetration, the soil around the penetrometer underwent a significant loading and unloading process, and the principal stress direction took rotation. The farther away from the penetrometer, the slower rotation speed of the principal stress, and the smaller the final rotation angle. The averaged pure rotation rate (APR) at different depths changed in a qualitatively identical way, but the APR maximum values increased gradually with the depth of soil. The cone penetration breaks the bond between the soil particles. There are two main kinds of bond failure modes, tension-shear breakage and compression-shear breakage, and the number of tensile-shear breakage is much more than that of compression-shear breakage.

Soil-water characteristic curve (SWCC) plays an important role in the study of unsaturated soils, and the research of prediction method of SWCC has a great scientific significance. The nuclear magnetic resonance (NMR) tests and SWCC experiments of Wuhan clay samples with different dry densities were carried out; Fredlund-Xing model was used to analyze SWCC experimental data and the complete SWCCs of total matric suction range were obtained. The theoretical relationship formula between matric suction and T2 value was established combining KST model and Young-Laplace theory. Based on the SWCC and NMR experimental results, the presented theoretical formula was tested. However, the test results show that this theoretical formula cannot effectively and simply describe the relationship between matric suction and T2 value. Through conversion and comparing analysis of SWCC and NMR experimental results, an empirical formula between matric suction and T2 value was established and testified to be rational. On this basis, a method was put forward to predict the SWCCs of unsaturated clay with different initial dry densities through NMR technique

In this study, dynamic pressure response of base structure of Fu Yingzi tunnel along Zhangjiakou to Tangshan was simulated through field large-scale excitation test under the class V rock conditions when the heavy haul trains with axle weights of 250 kN, 270 kN, and 300 kN passed. Dynamic pressure attenuation values at measuring points were presented when axle weight was reduced. According to the earth pressure time history chart of field measurement, horizontal and vertical distributions of dynamic pressure of the measuring points in each structural plane were analyzed. The results indicate that dynamic pressure of each measuring point in heavy haul line of double line tunnel is greater than passenger train line, and the dynamic pressure of the corresponding measuring points below the railway track under different axle load of heavy haul trains is greater than other measuring points. Stress concentration is found to appear in the structure and surrounding rock at corresponding heave haul position, which is more likely to cause the structure to crack or hole generation in wall rock under the condition of broken rock compared with the test results on the surface of class IV grade surrounding rock. Train load attenuates drastically due to cushioning of the bed structure and inverted arch filling as it vertically transfers, but there is only small attenuation in inverted arch. Current designed basement structure thickness cannot meet the demand of lifting axle load.

How to effectively assess the system reliability of slope containing unlimited number of potential slip surfaces and how to accurately identify key slip surfaces significantly contributing to slope failure are critical questions in slope engineering. To address these questions, this study proposed an approach to assess slope system reliability and to identify the representative slip surfaces based on generalized subset simulation (GSS). The equations for the evaluation of slope and quantification of the relative contribution of individual slip surface to slope are derived in this study. Using the reliability analysis results of GSS, the independent representative slip surfaces (RSSs) are identified from possible slip surfaces using probabilistic network evaluation technique (PNET). Finally, the proposed approach is verified using examples of a two-layered cohesive slope and a case of Congress Street Cut in Chicago. Results show that the proposed approach provides proper estimates of failure probabilities of individual slip surfaces and the slope system simultaneously by a single GSS run. This can avoid repeated simulations for each failure mode. The proposed approach is more efficiency than Monte Carlo simulation (MCS) especially for failure modes at low probability level and simultaneously overcome the limitations of subset simulation (SS). The RSSs are selected using PNET after obtaining the failure probabilities of the slope system and its corresponding potential slip surfaces. The slope system can be represented effectively by the RSSs identified in this study. The proposed approach effectively avoids the dependence of the accuracy of system failure probability on the number of RSS and the accuracy of failure probability of RSS.

During the classification of surrounding rock in tunnel engineering, different results are normally obtained by the same evaluation system and classification standard. It is caused by the uncertainty of rock parameters and the discreteness and randomicity of evaluation indices, which are resulted from the instrument error and artificial operation in the testing process. Especially in the sub-classification of surrounding rock, evaluation results usually have poor robustness, even grade skipping. Hence, it is necessary to investigate the reliability of surrounding rock classification. Based on the handbook of Engineering Rock Classification Standards, we analysed the probability distribution function of rock strength and rock mass integrity index. The system reliability analysis theory was also introduced to build the function of different surrounding rock grades. Moreover, Monte Carlo method was applied to calculate the reliable probability belonging to different evaluation levels. Finally, the reliability analysis method was proposed for the sub-classification of surrounding rock, according to the basic quality (BQ) method. Since this method takes into account the uncertainty and discreteness of the evaluation indices sufficiently, the calculated reliability indices can make a more steady evaluation result. Besides, the reliable probability can reflect the dispersion degree of working face information of the developed fault fracture zone and the weak interlayer to a certain extent. This method was applied to the surrounding rock sub-classification of Laohushan tunnel in Jinan belt highway. It was found that the evaluation results were consistent with the actual levels of surrounding rock. Besides, the continuity of reliability indices at different surrounding rock levels can quantitatively express the changing conditions of geological attributes during the tunnel construction. Therefore, this study provides data support for the transformation of construction method and the optimisation of supporting parameters. Research results have significant references for surrounding rock sub-classification in large-cross-section tunnels.

When using tunnel boring machine (TBM) to excavate a tunnel, engineers are hindered to observe the state of surrounding rock by the cutter-head and shield of TBM. Under this circumstance, the state of rock muck can be used to predict and evaluate rock mass conditions as well. The strength of blocky fragments in TBM muck can be obtained by the point load test. However, the relationship between blocky fragments and the intact surrounding rock has been unknown. In this study, point load tests are performed on the blocky fragments selected from the Yinsong Water Supply Project. Meanwhile, rock core samples are taken by drilling at corresponding locations of blocky fragments. Their strengths are also measured by point load tests, and then compared with uniaxial compressive strengths. In addition, geological conditions, the size of specimens and the breaking state are also recorded. The results show that the point load strength of blocky fragments reduces to 63.25% of core samples from side walls. It is found that the reducing ratio is greater when rock mass is more complete. The conversion factor from point load strength to uniaxial compressive strength for limestone is 25.3, whereas the conversion factor from the point load strength of blocky fragments to uniaxial compressive strength is 42.1. The breaking force depends on the breaking surface and oversize blocky fragments are not suitable for testing the strength. In conclusion, this study provides a useful method and basis for the rapid acquisition of rock strength in TBM tunnelling.

Groundwater distribution and determination of rock mass permeability parameters are very important in the process of underground engineering construction. Based on analysis of the objective function of underdetermined problem, and through the use of fitness function and the principle of geological statistics and variation function optimization, we obtain a class of hydrogeological problems' penalty function for solving formation permeability coefficient, which provides an optimized evaluation criterion to solve the problem of inverse analysis of underdetermined problem. Also, a mathematical model of optimized back analysis combined with particle swarm optimization (PSO) algorithm is established, and a new method to evaluate formation permeability coefficient is proposed using back analysis of the drilling head change from packer permeability test. At Nanjing Shang Yuanmen subway station, we conduct packer permeability test to get drilling head height on-site, and obtain permeability coefficient of the layer by using back analysis numerical calculation to optimize the penalty function. Detection of the resistivity cross-hole CT is conducted to verify the accuracy of the method, and drilling in-situ permeability tests verify the accuracy of the method to obtain numerical results. The results show that calculation accuracy reached 90% after analysis optimization by a penalty function, which proves that the method helps in getting comprehensive formation hydrogeological information. The method is important for the follow-up governance, which is expected to have certain reference significance for similar projects.

This paper is to investigate the calculation methods for the soil vertical deformation induced by constructing the ground penetrating shield tunnel. Lin’s formula was also corrected by considering the included angle ? between the axis of the tunnel and the horizontal plane (namely the variation of the buried depth of tunnel). A new formula of soil vertical deformation was derived by combining with bulkhead additive thrust, the friction force between shield and soil, additional grouting force and soil loss. Through analysing the examples, it was found that the results obtained from the new method were entirely different from those of Lin’s formula in a shallowly embedded tunnel. Especially, the results of the ground uplift before the excavation face and the settlement after the excavation face were both larger by the new method. When the shield tunnel excavated upward with the increase of ?, an upwards trend could be seen from the vertical deformation curve of longitudinal soil, which was induced by bulkhead additive thrust, friction force and soil loss. While in the same situation, a downward trend was found from the deformation curve of soil, which was induced by additional grouting pressure. In addition, the maximum value of the surface settlement reduced, whereas the range of the transverse surface settlement trough increased.

To avoid the damage of surrounding rock, pre-stressed anchor cables are installed in high side-wall and cavern intersection area to control the large deformation of surrounding rock. However, research is seldom performed on estimating the pre-load of cables in underground powerhouse with high geo-stress and low strength-stress ratio in the previous literature. Jinping-I hydro-power station is recognised as a typical high geo-stress underground powerhouse. Its cavern exhibits remarkable time-dependent deformation during excavation, which may result in the continuous increase of the inner force of the anchor cable and bars, even the strength failure. To determine an appropriate pre-load, a procedure is proposed for calculating the time-dependent released load of surrounding rock, based on their deformation characteristics. At the same time, a formula is deduced to calculate the pre-load factor of pre-stressed anchor cable. This established formula considers the influences of geo-stress, anchorage force and installation time on the pre-load factor. Comparing the actual values with the theoretical values of pre-load factor of Jinping-I underground powerhouse, the result indicates that the established calculation formula for the pre-load factor is reasonable and can be referenced in practical underground engineering.

Translational landslides are widely spread in red layers area in Sichuan Basin. This type of landslide develops in sub-horizontal bedrock composed of sandstone and mudstone, normally with the dip angle of 3 to 5 degrees, 10 degrees at most. Based on geo-mechanical model and sliding process analysis, with the application of the principle of energy conservation, we derived the theoretical sliding distance calculation formula of translational landslide. Three groups of physical simulation tests were established. Group test Ⅰ simulated the sliding distance under the water pressure generated under different conditions of widths and different water heads. Then the theoretical formula was tested by comparing the calculated results and group Ⅰ physical simulation test results. By comparison, the overall distribution shape of uplift pressure under the bottom of sliding body was close to rectangle with very few groundwater penetration points along the shear outlet. To verify the theoretical formula, group tests Ⅱ and Ⅲ were established thereafter. Each group of tests fixed the width and water head respectively. Meanwhile the shear outlet of physical model was set to two types totally blocked and totally seeping water. The results of three groups of physical simulation tests verified the reliability and applicability of theoretical formula. Applying the theoretical formula to Shizishan landslide to calculate the sliding distance, and the calculation results showed the good applicability of the method. The research has some guidance and practical value to prevention and mitigation of translational landslides.

Excavations of large-scale hard-rock underground caverns with high stress often result in rib spalling and collapse of surrounding rock, threatening the safety of on-site builders and equipment. From the perspective of reducing the risk of local failure of surrounding rock, framing excavation method was adopted in the fourth floor of the right transformer chamber of the Baihetan hydropower station. In order to study the microseismic (MS) characteristics and stability, MS monitoring was carried out in the right transformer chamber. The results show that: By the excavation of upstream rock, the number of MS events and energy release were greater than the downstream rock. In the process of working face advancing, the number of MS events and the energy increased, which was in accordance with the overlay abutment pressure. When the upstream rock was excavated, the excavation unloading induced seismic events and the cluster phenomenon. In contrast, the excavation of downstream rock only induced few MS events. The Schmidt number, the activity rate, the cumulative volume, and the energy index were combined to evaluate the stability of the study area. The study found that compared with the strain hardening phase, the rock mass in the stage of strain softening has a greater risk of unstable deformation and failure. The results can provide valuable reference to the optimization of construction scheme of the Baihetan hydropower station. At the same time, the results can also provide reference to other similar projects of construction.

Owing to the excellent shock absorption property of the foamed concrete, it can be used as a seismic isolation material for tunnels. In this study, a series of compression tests was conducted to study the effects of density, confining pressure and strain rate on mechanical properties of the foamed concrete. As the density increased, the failure mode of foamed concrete gradually changed from cell wall buckling to shear failure under the uniaxial compression condition. Besides, the strain-softening and material brittleness became more apparent after the peak. In addition, the strength and ductility of the foamed concrete increased with increasing confining pressure. Moreover, the property of foamed concrete transformed from a strain-softening to strain-hardening material. When strain rate increased in the range of medium strain rate (10?5/s～10?3/s), the strength of foamed concrete showed an exponential growth, and the residual stress in plastic range increased obviously. It was found that the effect of strain rate on the residual stress was greater with the increase of plastic strain. Based on the testing results, a new constitutive model was initially proposed for the foamed concrete. Then the numerical method was performed to investigate the effects of the shear modulus of isolation material, the thickness of isolation layer, and the properties of isolation interface on the isolation of Galongla tunnel. Therefore, the results can provide a helpful reference for the design of the related cushioning layer in tunnels with high intensity.

Structural marine soft soil layer is widely distributed in coastal region, which brings lots of potential disasters. Soft soil subgrade in the Jiangsu provincial section of an expressway is studied in this research. A new method, to obtain constitutive parameters of undisturbed soil of Shanghai model by consolidation test results of compressed soil is proposed. The subgrade settlement is simulated by soil-water coupled FEM software. The effect of soil structure, permeability and thickness of soft soil layer on subgrade settlement are analyzed. The results show that the stronger the initial soil structure is, the larger pore pressure and subgrade settlement are in loading stage. With the gradual destruction of soil structure after completion of construction, it is positively correlated with the amount of layer compression and soil structural parameters. The permeability affects settlement and settling time as well. The lower the permeability is, the larger total subgrade settlement and more proportion of post-construction settlement are. Thickness of the soft soil can also affect settlement and settling time. The thicker the soil layer is, the larger subgrade settlement and longer settling time are.

The study on compaction characteristics of fractured rock mass is considered as one of the basic work of underground engineering in coal mines. Since fractured rock mass exists in the hidden and dangerous environment, laboratory experiments and numerical simulations are normally applied to determine the compaction characteristics indirectly. To construct three-dimensional (3D) model of fractured rock mass, this paper presented a novel method which was based on the 3D Voronoi numerical model of intact rock mass. In this model, the porosity was pre-determined and the structure of fractured rock mass was retrieved by randomly removing the blocks in the intact rock mass. Moreover, the Weibull distribution was introduced to describe mechanical parameters of joints. In addition, uniaxial compression experiments were conducted to investigate the compaction characteristics. Compared with existing theoretical and experimental results, this method demonstrated good agreements with them. It is proved that this parametric method can reflect the fragmentation, bulking and compaction characteristics of fractured rock mass more accurately. Therefore, this study provides a new and effective method for the safety control of underground engineering in coal mines, which can be widely used in the gravel experiments.

Due to the inconsistent stratigraphic sequence of the stratum in the borehole sample, it is difficult to accurately determine the topological relation of each stratum in 3D stratum model based on borehole data. Hence, one of the key technologies in the 3D geological model is to make full use of the borehole data. Based on the theory of geological interpretation method, the stratigraphic sequences were confirmed automatically through the frequency of the stratum appearing at the upper location of all boreholes in the region. By using the surface modelling method, the recursive method of sub-drilling was first introduced to establish the 3D stratum model layer-by-layer from bottom to top. Meanwhile, this method can also adapt to the complex geological formation, such as strata pinch-out, stratigraphic overlap, and lens. The proposed 3D geological modelling algorithm with the advantages of precise geoscience meaning, good robustness, high efficiency and reliable operability, was already implemented into TJSG software and applied to some practical 3D geological modelling projects successfully. The research results indicate that the proposed algorithm can not only fully use all the borehole information to create a 3D geological model, but also has strong adaptability for complicated stratum.

Horizontal deformation resulted from excessive groundwater exploitation has received attention in recent years because of earth fissures and technological development of deformation monitoring, such as GPS and InSAR. However, traditional uncoupling 1D model of land subsidence cannot simulate horizontal deformation. Though Biot model is able to simulate horizontal deformation, using this model is not appropriate for numerically solving the problem of regional land subsidence because the size of discrete model and time step is limited due to long computation time and ill-conditioned linear system matrix. To overcome inherent weaknesses above, uncoupling 3D mathematical model of land subsidence is developed by combining the benefits of uncoupling 1D model and Biot model. Uncoupling 3D mathematical model consists of groundwater flow and deformation equation. Both equations are coupled with parameters (i.e., Young modulus, Poisson ratio and specific water storage). Groundwater flow equation initially computes 3D flow field (equivalent fluid pore pressures) and then the deformation equation calculates 3D deformation based on the known flow field. The derivation of uncoupling 3D mathematical model shows that Biot model can be simplified to uncoupling 3D model based on the assumption that the total normal stress does not change in the process of groundwater flow; uncoupling 3D model can be simplified to uncoupling 1D model based on the assumption that the radial displacements vanish. Meanwhile, numerical experiment shows that uncoupling 3D model can provide deformation results similar to Biot model with much less running time compared with Biot model. Therefore, uncoupling 3D model can be considered as an alternative model to Biot model and an improved model for uncoupling 1D land subsidence model to simulate 3D regional land subsidence.

Spatial variabilities of soil properties are usually modeled by stationary or weakly stationary random fields in slope reliability analysis. However, abundant in-situ data demonstrates that the means and standard deviations of soil properties such as undrained shear strength change with the soil depth. Thus, it is important to develop non-stationary random field models and approaches to model spatial variations of soil properties. To tackle the uncertainties of the trend component and fluctuation of component in existing models, this paper proposes an effective non-stationary random field model of undrained shear strength. The approach for modeling the two-dimensional non-stationary random fields of soil properties is also presented. Thereafter, the effectiveness of the proposed model is illustrated by comparisons with the existing non-stationary random field models and stationary random field models of soil properties systematically. A clay slope under undrained conditions is investigated to explore the effect of the non-stationary distribution characteristics of undrained shear strength along the depth on the slope reliability. The results indicate that the proposed model can simulate the uncertainties of the trend component and fluctuating component separately. Both the means and standard deviations of soil properties increase with the depth. It provides an effective means for characterizing the non-stationary distribution characteristics of soil properties. Compared with the slope reliability analysis results underlying the non-stationary random fields, it will lead to an underestimation of the probability of failure and dangerous slope engineering designs when the commonly-used stationary random fields are utilized to model the spatial variability of soil properties.

Deep rock is characterised by high stress and large deformation in underground engineering. It is of great significance to investigate its mechanical properties by conducting the triaxial tests under high confining pressure and considering a residual strength. The distinct element method (DEM) is an essential numerical method for analysing the mechanical properties of rock. However, there have been many challenges in quantitatively simulating the rock by DEM in the triaxial tests. The reason is that it is challenging to quantitatively reproduce the complete stress-strain curves of rock using DEM. By considering the residual strength of rock, the triaxial compression tests on deep sandstone were quantitatively simulated by DEM using the improved three-dimensional (3D) bond model incorporating rolling and twisting resistance. The results show that this study successfully achieves the quantitative reproducing of the complete stress-strain curves. The large peak/residual internal friction angle and nonlinear strength envelope of rock are obtained as well. Meanwhile, this study overcomes three prominent challenges of the classic bonded-particle model (BPM). The relationships among the peak/residual internal friction angle, cohesion and microscopic parameters of DEM are discussed through a large number of simulations. At the same time, the high computational efficiency is proved using the improved model, which can satisfactorily meet the requirements of the 3D simulation for conventional laboratory tests.

The failure of geological discontinuities often plays a significant role in controlling the failure of a rockslide. The second order work criterion can be used to analyse physical instabilities, except flutter instabilities. Hence, this study was to investigate the instabilities of geological discontinuities using the second order work criterion. First, regarding the instability of the major geological discontinuity, the advantage of this failure criterion was explained in mathematical approach; Second, the criterion was implemented into the FLAC finite-difference calculation software, and a two-dimensional numerical analysis was carried out to predict the occurrence of the rockslide in Nanfen Iron Mine, according to the corresponding monitoring data. Rockslides can be considered as a quasi-static to the dynamic transition, in which the sudden burst of kinetic energy indicates the occurrence of quasi-static-dynamic transition. Moreover, there is a direct relationship between the second order work and the second order kinetic energy. Therefore, the kinetic energy evolution curve can be estimated, which can prove the validity of the second-order work on identifying the failure of the rockslide.

Stone retaining wall is constructed to resist the lateral earth pressure. Its characteristic is regarded as a discontinuous medium, whereas the characteristic of the soil behind the wall is a typical continuous medium. In this paper, the ultimate bearing capacity of masonry retaining wall was studied by combining the lower bound theory, the mixed numerical discretisation and linear programming. Firstly, soil mass was discretised by triangular finite elements to simulate its continuum mechanics characteristics, while the masonry wall was discretised by rigid block elements to simulate its non-continuum mechanics characteristics. Thus, constraint conditions of the statically admissible stress field were both proposed for soil and the masonry retaining wall. Meanwhile, constraint conditions were also proposed for interfaces between finite elements and block elements. By considering the overload coefficient as the objective function, the linear programming model was established for calculating the ultimate bearing capacity of masonry retaining wall. In addition, the interior point algorithm was employed to solve the problems of linear mathematical programming. At last, the ultimate load (or safety factor) of slope and its corresponding stress field were obtained directly. Moreover, three typical examples were conducted to validate the proposed method. Therefore, this study successfully introduced the mixed numerical discretisation into the limit analysis.

Water saturation in rock significantly affects its mechanical parameters. However, the existing methods for preparing rock specimens with different water saturation levels have some common problems, such as low accuracy, long time, and uneven distribution of water. This paper proposed a rapid and efficient procedure to prepare rock samples with different water saturation degrees based on chemical thermodynamic principles. This method can control water saturation of rock specimens precisely by adjusting the humidity of environment with a negative pressure, which governs the fugacity of water in rock samples. Since the fugacity of water vapour in constant humidity environment is identical, the water vapour and the content water in rock samples would form a fugacity difference when the fugacity of the water vapour and the content water are different. Content water in rock samples would escape from rock samples when the fugacity of water vapour in the environment is less than that of content water in rock specimens. Otherwise, water vapour would infiltrate into rock samples. Eventually, rock specimens with certain water saturation levels are obtained when the fugacity of content water in rock specimens and water vapour in the environment are equal. Rock specimens were prepared in a negative pressure environment. Air in the porosity of rock specimens would decrease and the seepage velocity of water vapour in rock samples increases in this kind of environment, which can shorten the time of rock specimens to reach the steady state. The fugacity of content water in every component of rock specimen is equal when rock specimen reaches the steady state, according to vapour-liquid equilibrium principles. Therefore, the distribution of water in rock specimens is even.

This study introduces the main function, technical index and instrument composition of a developed rock multi-functional shear test system in detail. A series of mechanical experiments is carried out by using the testing system. This system mainly consists of three parts: testing device, measurement system and control system. The maximum normal tensile stress is 40 MPa, and both the maximum normal compressive stress and horizontal shearing stress are 120 MPa. The specimen size is 50 mm?50 mm?50 mm. Multiple mechanical experiments are conducted on granite, including direct tension test, direct shear test, tension-shear test and compression-shear test. The results show that brittle fracture of the specimen occurrs in the direct tensile tests. The acoustic emission (AE) signal instantly reaches the peak and the fracture surface exhibits the characteristics of tensile fracture. In the direct shear test, the specimen is destroyed many times. The AE signal suddenly increases at the moment of destruction, and the fracture surface shows the characteristics of shear fracture. In the tension-shear test, the shear strength of the specimen decreases significantly under the tensile stress. The AE signal is obvious at the failure stage, and the fracture surface exhibited both tensile and shear fracture characteristics. The results of the above mechanical experiments indicate that the developed multi-functional shear test system can successfully carry out a variety of mechanical experiments and provide a new testing method to further study shearing properties of rocks.

The micro-scale mechanical behavior of granular soils (e.g., particle movement and particle breakage, etc.) governs their macro-scale stress and strain behavior, such as strain localization and stress hardening, etc. To study the micro-scale mechanical behavior of granular soils, a mini-triaxial apparatus is developed in this paper. The axial loading system of the apparatus is composed of a servo-controlled stepping motor and turbine-driven reducer. The confining pressure is provided by a GDS pressure generator, and the chamber is fabricated with a highly transparent material with high strength and low density. In conjunction with an X-ray micro CT image processing and analysis techniques, this apparatus can be used for non-destructive detection of the micro-scale characteristics of a mini dry sample (i.e., 8 mm in diameter and 16 mm in height) under triaxial shearing. A triaxial test of a Leighton Buzzard sand (LBS) sample with an initial grading of 0.60-1.18 mm under a confining pressure of 1.5 MPa is carried out using this apparatus. The results show a reasonable stress-strain curve and CT images with easily distinguishable features, which demonstrates that the apparatus can be used for testing of the micro-scale mechanical behavior of granular soils.

Due to the limit of complex terrain, the structural plane of rock mass is difficult to measure in the geological survey of high and steep slopes. Hence, a new technique is in urgent need to achieve all-around precise measurement. In recent years, the technology of the low-altitude and low-speed small unmanned aerial vehicle (UAV) has been developed rapidly. In this study, three-dimensional (3D) reconstruction of the target was performed using the small UAV with a single-lens reflex camera, based on the patch-based multiview stereo (PMVS) and structure from motion (SfM). By means of field experiments and data post-processing, an application method of the UAV oblique photogrammetry was proposed for the geological survey of the high and steep slope. The preliminary conclusions were also obtained. The valid terrain data was successfully collected by the small multicopter UAV with the single-lens reflex camera. During the process of analysing 3D point clouds, the parameter of the structural surface was extracted by the plane fitting algorithm based on the least square method. The structure plane of rock mass was further drew into the polar-radiation equatorial-plane projection map. Thus, the digital measurement of rock mass in high and steep slope was achieved in this study.

Taking the rock porosity as the focal point of rock damage analysis, this study is to develop a capillary infiltration technique for non-destructive quantitative analysis of the geotechnical damage. Moreover, variations of the capillary infiltration rate, infiltration area and depth of infiltration are investigated by changing the porosity in capillary infiltration tests. The capillary bundle model is proposed and the mechanism of capillary infiltration technique is also analysed in depth. Compared with the existing damage and non-repeatable measurement technologies, this technology is easy to operate and low cost. In particular, for a rock with specific damage, we can establish a relationship between titration parameters and mechanical parameters. In addition, the capillary model 1, the cylindrical model 2 and the spherical model are put forward, respectively, by considering the effect of capillary infiltration on the lower part of the capillary body in the capillary bundle model. The depth of the liquid infiltration in the lower porosity L is estimated from the upper liquid column descending height and the surface of infiltration radius a. According to the infiltration parameters of the actual pore soil, the validation of the above model is verified. In addition, the result indicates that the spherical infiltration model can better simulate the infiltration depth of the liquid during the process of capillary infiltration. Based on the capillary bundle model, the porosity calculation formulas of three models are separately deduced by the infiltration rate, which is the relationship between the infiltration depth and the infiltration rate. The calculated porosity of the spherical cap model is examined with the measured porosity, and its error is less than 10%. Therefore, the results calculated by this porosity formula are reasonable and effective for general accuracy requirements.

A cyclic loading device was designed suitable for laboratory model test. The device, consisting of the motion system, driving system and reaction frame system, has the advantages of stability, low cost in manufacturing, adjustable load frequency and alterable amplitude of loads. The cyclic loading device is used to investigate the dynamic response and cyclic behavior of monopiles installed in the laboratory test box. The test results indicate that the similarity between measured load and theoretical load is about 91.3%, indicating that the load exported from the device is very stable and accurate. The accumulated deformation of monopiles in the soft clay increases at the beginning then gradually tends to be stable. Soil behind the pile is separated from monopiles and the cracks are formed in the surface soil surrounding the pile. The maximum moment along monopiles increases with the number of cyclic loads at the beginning, then tends to be stable after about 1 000 cyclic loads. The location of maximum moment of monopiles is about 4-5 times of pile diameter under mudline. It shows that the design of the loading device is rational, and suitable and reliable for the model test.