Advances in Earth Structures: Research to Practice

Great progress has been made in developing three-dimensional (3D) slope stability analysis methods based on the `method of columns' since Duncan summarized the main research outcomes in this area. In the limit equilibrium framework, Chen et al. presented an extension of the 2D Spencer's method that satisfies the force equilibrium conditions in all the three co-ordinates and the moment equilibrium condition in the main axis of rotation. In the upper-bound limit analysis area, Chen et al. gave a generalized approach that is actually an extension of the 2D Sarma's method. Firstly, this paper presents recent advances in 3D slope stability and a number of case studies that have solved the key slope stability issues in China's hydropower projects. Secondly, recent advances in development/refinement of 3D slope stability analysis methods are reviewed and summarized in this paper. Finally, a forward look is highlighted at the end of the paper, including an overview of further works that are needed for more research and a few geotechnical engineering areas in which the use of 3D stability analysis could be promising.

The design objective of geosynthetic reinforced earth structures is to determine the economical layout and strength of reinforcement rendering a certain criterion. Outlined are two complementary approaches that satisfy this objective. The approaches are based on limit equilibrium analysis. The analysis considers the interaction of soil-geosynthetic and the interaction among reinforcement layers. The end result ensures sufficient length and strength of reinforcement. It also considers the required strength at its front end at the connection to the facing units. One approach produces the required tensile resistance of the geosynthetics so as to render the same factor of safety everywhere. This approach yields the "demand" aspect in design and therefore is considered fundamental to soil reinforcement. The second approach considers the "supply" or the existing strength of reinforcement and renders the safety factor everywhere. This approach is the practical one as it matches the supply with the demand via safety factors. While such an approach is not novel, this paper demonstrates how it can be used effectively and objectively using color-coded safety maps to optimize the reinforcement length and strength.

Experimental research was performed to evaluate the use of small-diameter, structural pile elements for stabilizing slopes. Full-scale lateral load tests were performed on individual drilled and grouted pile elements, in which the piles were installed through a shear box into stable soil and then loaded by uniform lateral translation of soil. Instrumentation of the shear boxes and pile reinforcement indicated the load development along the piles and the pile response to the physically-imposed boundary conditions. The load test analysis was then conducted to evaluate an existing analysis method for its applicability to isolated, slender piles. Experimental testing results validated the lateral response analysis method for piles subject to lateral soil movement and helped to characterize the relationship between lateral soil movement and pile behavior. Using the analysis method, a new approach is described for evaluating the stability of pile-reinforced slopes considering both strength and service limit states. Allowable slope movements for reinforced slopes constitute a practical design parameter for movement sensitive structures like buried utilities, etc.

Slope failures resulting from seismic events are unfortunate yet relatively common occurrences. Subsurface deformation patterns in failing slopes can be studied using physical models, but it requires relatively small and non-intrusive sensing instrumentation. Time domain reflectometry (TDR) technology, which has been successfully applied to field monitoring of unstable slopes during the past two decades, offers a diverse monitoring tool for physical models, particularly for tracking localized shear deformations with an embedded sensor array. This paper presents the setup and results of 1-g slope failure model tests to demonstrate the capability of TDR to capture localized shear deformation. Two prototype coaxial cables, one grouted and the other ungrouted, were installed full depth as TDR sensors. Experimental results show that grouted and ungrouted sensors accurately locate the failure surface after sliding commences. For the same soil conditions, a grouted sensor produces a higher rate of reflection during sliding than an ungrouted sensor because the interaction length across the shear zone is minimized.

The safety factors of the current design codes relating to soil structures don't always provide absolute and objective indices. In order to investigate the cause of this limitation, tilting failure tests of sandy soil model slopes and Bishop's simplified method were performed under three confining conditions at the toe of the slope. We introduced new strength parameters, c_γ and φ_γ, which are functions of shear strain γ. The limit equilibrium analysis using the values of c_γ and φ_γ corresponding to shear strains distributed in the shear band can explain well the actual failure mechanism of a model slope in every case of the three confining conditions. These results strongly suggest that the limitation of the safety factors stated in the first sentence might result from such an over-simplified failure mechanism as a priori introduction of a slip surface, which means "a simultaneous failure" all over the slip surface.

Establishing the rheology parameters of rock masses is one of the important tasks for the long-term stability of the left bank high slope at Longtan hydropower station in China. Existing back analysis methods are not sufficient to provide the necessary accuracy of rheology parameters. The new intelligent displacement back analysis method proposed in this paper is a combination of a neural network, a genetic algorithm, and FLAC^3D—the evolutionary neural network finite difference method. The rheology parameters of rock masses can be accurately obtained. With these parameters, a numerical simulation is conducted, the results of which indicate that the measured displacements and the calculated displacements are similar in value and trend of deformation.

We show the two types of a progressive failure by strain energy, U: strain energy released and accumulated propagations. We also have clarified the failure mechanism of strain energy accumulated propagation by model tests and theoretical examination of a simple model. As a result, we have been able to show that the external force condition classifies the strain energy accumulated propagation into the two types: forced and self-excited propagations. Only forced propagation occurs in soft ground, but both forced and self-excited propagations occur in hard ground. Verification tests have shown the validity of most of the examination results.

A simplified methodology is proposed for the stability analysis of slopes reinforced with one row of piles. To account for the presence of piles, an existing method which is based on the theory of plasticity is used to determine the lateral forces acting on the piles. Then, the kinematic approach of limit analysis is used to analyze the stability of the slope. The rate of external work due to the pile force is added to the energy equation and expressions are derived to calculate the safety factor of slope reinforced with piles. Using the proposed method, a study is carried out to illustrate the effect of piles on slope stability. Finally, the most suitable location of piles within the slope and the effect of pile spacing and diameter on the safety factor are discussed.

This paper presents a new method, which could consider progressive failure of reinforced slopes subjected to seismic loads within the framework of limit equilibrium method and pseudo-static approach. The failure surface has been modeled employing a generalized log spiral and the sliding block was divided into horizontal slices. Two types of safety factors have been used. One is for the overall stability, defined for the whole sliding wedge, and the other represents the local stability of each slice. Peak strength parameters have been used to determine the most critical slip surface and the length of reinforcements. Then the analysis is repeated applying the peak strength parameters for slices, with safety factor more than unity and residual strength parameters for slices, with safety factor less than unity. At last the required reinforcement strength required to maintain the stability of slope is determined. The results of the presented method have been compared with published data. It has been concluded that Horizontal Slice Method (HSM) could properly model progressive failure in seismic condition and represents more realistic results.

The equality distribution, Poisson distribution and fractal distribution were calculated for the data about the time distribution of landslides in the Three Gorges Region. The fractal of landslides was shown in the narrow spectrum near the natural cycle. According to different size ranks, the approach of physics fractal was proposed to divide the landslides into 6 subsets and study their dimensions (D), scale spectrums, and natural cycles. They indicated the similarity and hierarchy of landslide occurrence in the fractal. The scale spectrums for whole landslides were 2.7–4.2 and for size ranks were 1.98–5.84.

In the granitic mountains of West China, the spectacular scenery that has developed under arid conditions includes the presence of sand ramps lying at the toe slopes. The analysis of ramp geomorphology and sedimentology, and rock slopes above the ramps indicates that the ramps were formed by deposits of sand hillslopes after sliding. The process of a typical sliding sand hillslope (SSH) involves three major events: formation of sand due to weathering of rocks in upland areas and slopes, transportation of sand along the hillslope, and accumulation of sand at the toe of the slope. SSH commonly exists in the granitic mountains of Tibet and Xinjiang, however, it is one of the landforms rarely reported from mountain environments. This paper presents engineering and geological characteristics of SSHs based on field investigations and laboratory tests. This study investigated major damages of SSH and possible causes for intensive occurrence of SSH in this specific area. At the end of this paper, remedial measures to protect highways in this area are proposed.

The objective of this paper is to present the results of a geotechnical monitoring of a landslide movement prior, during, and after stabilization and repair of the slope in Orange County, California. The slope is a combination of graded slopes and retaining walls descending from a freeway to a railroad alignment and finally down to a channel. An investigation program consisting of geotechnical borings, field and laboratory testing, and installation and monitoring of inclinometers and piezometers were performed. The purpose of the investigation was to evaluate the causes of the instability and to recommend alternative stabilization methods. The final method adopted for construction consisted of a retaining system of mini-piles and tie-back anchors.

The shear strength of slip soils is important for the stability analysis of landslides. Most of the previous research is mainly focused on the strength characteristics of finer particles; however, little attention is paid to the strength behavior of slip soils containing coarse particles. This paper presents a series of shear tests on slip soils sampled from the adit of Xietan landslide in the Three Gorges Reservoir areas. The shear tests include ring shear, multi-reversal direct shear (both small and large), triaxial shear (CD/CU), with the proportions of coarse particles ranges from 0% to 40%. Results indicate that both the gravel proportions and its size have an effect on the shear strength behavior. The shear strength parameters of the slip soils of Xietan landslide are presented for analyzing the stability of landslide.

The convention approach to assessing slope stability is through the calculation of safety factor. However, shortcoming of this approach is that uncertainties in slope system are not explicitly reflected in safety factor. Reliability theory can consider uncertainties in slope system. However, reliability theory is complicated and short of experience. Therefore it is necessity to establish dual index system of slope stability coupling safety factor and reliability theory. Most likely safety factor (mean-value safety factor) was calculated with average value of parameters, and reliability of slope was calculated with variability of safety factors. It was assumed that safety factors obey normal distribution or lognormal distribution, a series of most likely safety factors and failure probabilities of slope P_f were calculated with different coefficient of variation. Most likely safety factors were multiplied by corresponding reliability (R=1–P_f), and most likely safety factors were reduced. From limit state of reduced safety factors, critical failure probabilities corressponding different most likely safety factors were calculated. According to different most likely safety factors and their failure probabilities, a divisional chart of slope stability was drawn, and then a dual index system of slope stability coupling safety factor and reliability theory was established.

The waste disposed in MSW (municipal solid waste) landfill consists of material in different shapes with various physical and chemical properties. The engineering properties of waste change with time and make it a challenging task to correctly predict and analyze the long-term deformation and stability of solid waste landfills. The strength of waste usually decreases with time and affects stability of landfill whereas most of the engineered earth fills are more stable with time. The stability of solid waste landfills is most likely jeopardized by excessive fill of waste when lack of proper management. A slope failure case of an old MSW landfill is introduced and analyzed with respect to different phases of filling and operation history to illustrate the stability hazard from uncalculated excessive fill. A parametric study on strength of waste and underlying native soil is also presented.

3S (Remote Sensing, Global Position System, and Geographical Information System) technology can be applied in regional landslide forecast. This paper discusses these new technologies and methods applied in landslide forecast. RS (Remote Sensing) technology can accurately provide source data of different periods in time. Depending on its strong function of spatial analysis and spatial database management, GIS (Geographical Information System) technology can pick up the basic factors of landslide quickly and efficiently. GPS (Global Position System) technology can provide convenient and exact means for landslide inspection. In this paper the authors used the Yunyang to Wushan segment of the Three Gorges Reservoir area as an example and analyzed the application of 3S technology in landslide forecast in detail.

A large high-speed landslide occurred in the Touzhai valley in Yunnan province of China on September 23, 1991. The slide mass, approximately 18 million meters cubed in volume, moved with an average velocity of 28 m/s. Eventually part of the slide mass plunged into the Pan-he river within 2–3 minutes. The slide mass moved about 3.65 km. Clastic basalt filled up the 40 m deep valley and formed the 250 m wide landslide deposits. Detailed investigation of the landslide showed that the slide could be divided into three related stages according to time and space. The stages were set-out, short-range and long-range. Each stage had different hydrokinetic phenomena and motion characteristics. The slide mass slid rapidly along the slip surface in the set-out stage, flew at a speed of over 54 m/s in the short-range stage and turned into the fragment flow in the long-range stage. This paper analyzed the mechanism of high-speed and long run out all over the course and came to several conclusions.

High slope stability problems widely exist in geotechnical engineering practice. It is very important to monitor deformation of high rock slopes, timely process and analyze the measured data to ensure safety of construction. A high rock slope at the left bank of Lijiaxia arch dam in China is selected here to demonstrate the utilization of rock deformation monitoring data to analyze and evaluate its stability. The least-square method was used to establish the progressive regression models of rock deformation based on space-time qualitative analyses of the monitoring data measured after the impounding of the reservoir and physical cause of formation analyses of rock deformation doubtful points. The results were utilized to analyze the influence of hydraulic pressure, temperature, and time on the high rock slope deformation. The following conclusions are obtained in this study: (1) the high rock slope slid downwards the riverbed during the reservoir impounding period, (2) the deformation of the rock slope gradually became stable after the impounding water level reached the normal storage level, and (3) Portion of the left bank still has a tendency to slide downwards slightly. Suggestions were made to enforce the safety monitoring.

A new design methodology was developed to analyze the stability of soil nailing and forces in soil nails. In soil nailing, nails were generally designed to be parallel to each other. A new way to separate soil and nails into slices was developed in this study, in which the failure wedge of soil and nails were divided into a number of slices parallel to nails. By the proposed method, both the safety factor of soil nailing and the forces in soil nails could be obtained respectively. The critical slip surface and the limit state of soil nailing were determined by the proposed method combining with modified genetic algorithm (GA) which was based on simple GA and merged with traditional optimal methods such as complex method and had the advantages as global convergence and fast convergence rate. This proposed method was validated by comparing the predictions with practical observations.

After the Three Gorges Dam is built, the reservoir will be impounded gradually untill the level rises to 100 m. As the need of flood prevention, the reservoir's water level will fluctuate between 145m and 175m. The intrinsic water-rock interaction circumstance and condition will be changed by the periodical rise and drawdown of water level, which may result in the instability of reservoir banks and slopes. The Baisuigou landslide is the secondary landslide of the Tanjiaping landslide, which is located at the new immigrant county of Badong County in Three Gorge region. It is important to ensure Baisuigou landslide. Based on a number of experimental studies and monitoring data, a 2D FEM model was developed. Under the fluctuation of different water levels, the deformation trends, the failure process of the Baisuigou landslide, and effect of prestressed anchor piles on controlling landslide are simulated by ANASYS. The research results show that due to the periodic water level fluctuations, the landslide stability will decrease and the landslide will happen at last. However, the landslide stability can be increased by the prestressed anchor piles. This research will be useful for predicting the stability of Baisuigou landslide in long periodic water-level fluctuation.

The objective of this paper is to find the permanent deformations of Tehri dam due to an earthquake of magnitude 8.5, the occurrence of which has a high probability in the region, and for an earthquake of magnitude 7.0, for which the dam has been currently designed. Instead of finite element analysis, five different empirical methods, Makdisi and Seed's method, Newmark's double integration method, Jansen's method, Swaisgood's method and Bureau's method have been utilized and their results compared to get a range of values within which the permanent deformation of the dam is estimated to lie. The present analyses show that the predicted deformations due to an earthquake of magnitude 7.0 do not compromise the safety of the dam. However, the deformations predicted for an earthquake of magnitude 8.5 are quite high. The maximum deformations are predicted by Makdisi and Seed's method while the minimum deformations are computed by Jansen's method.

Geomembrane, a low permeable synthetic material of relatively low elastic modulus and tensile strength, has been widely used as a barrier/containment system. Guidelines in the USA typically limit the allowable strain in geomembrane during service up to 5%. This paper addresses a townhouse development project in the USA where a layer of geomembrane was placed within the foundation system because of environmental considerations. The foundation system consisted of a load transfer platform with four layers of geogrid reinforcement over Vibro-Concrete Columns (VCC) in a triangular pattern. Due to the complexity of this problem, a finite difference method was used to compute the tension and strains in the geomembrane during construction and service. A linearly elastic-perfectly plastic model with Mohr-Coulomb failure criterion was employed to model the foundation soils, the load transfer platform fill, and VCC columns with a very high cohesion. The townhouse strip footing was modeled as a linearly elastic material. Geogrid elements were used for both geogrid and geomembrane sheets. Distributed loads of two different magnitudes were applied at locations of footing and floor respectively to simulate the footing load and the floor surcharge. The construction was modeled in six stages. The analysis showed that the geomembrane would survive during and after the construction of the geosynthetic-reinforced column-supported platforms when subjected to footing and floor surcharge loading.

Piled reinforced embankments have been increasingly used to construct highways on soft soils due to their rapid construction, low costs, and small total and differential settlements. The estimation of pile efficacy is important in design. Pile efficacy is obtained from an analytical model of piled reinforced embankments based on the principle of minimum potential energy, which considers the interaction among embankment fill, geotextile, pile (cap), and subsoil. The effects of pile spacing, embankment height, cap size, shear modulus of embankment fill, geotextile stiffness, and pile-subsoil stiffness ratio on pile efficacy are investigated in this study. The results show that pile efficacy is significantly affected by pile spacing, embankment height, cap size, and pile-subsoil stiffness ratio, while shear modulus of embankment fill and geotextile stiffness have lesser effects. The theoretical predictions compare reasonably well with field measurements.

Residual undrained shear strength, s_u, of granular soils is often estimated from back analysis of post-earthquake geometry of embankments. In these analyses soil behavior is usually assumed to be inherently isotropic although saturated sandy soils exhibit strong inherent anisotropy under undrained loading. This paper presents back-analyses of two case histories of undrained instability of embankments observed during the Magnitude (Japanese Meteorological Agency) 7.8 (M_JMA 7.8) Kushiro-Oki Earthquake of January 15, 1993 considering inherently anisotropic soil behavior. The residual undrained shear strengths obtained from these analyses were then compared with the corresponding estimates from conventional back-analyses which assume inherently isotropic soil behavior.

Rigid pile-supported embankments are widely used to construct highways, dams, storage tanks, and so forth, on soft soils. The interaction between the components of a piled embankment including the embankment fill, the pile, and the foundation soil is complicated. The pile, the embankment fill, and the soil are assumed to deform one-dimensionally. An approximate closed-form solution is obtained, which considers soil arching in the embankment fill, the negative skin friction along pile, and the settlement of the substratum. Studies on examples with different soil profiles and pile lengths indicate that the solution is consistent with those obtained from a rigorous finite element method (Plaxis). The stress concentration on the pile top, load distribution along pile, the settlement of the embankment, and the soil are investigated.

The stabilization of soft clay is one of the most important construction techniques in geotechnical engineering. Chemical admixture (lime and cement as chemical admixture) stabilization has been extensively used in both shallow and deep stabilization in order to improve inherent properties of soil, such as strength and deformation behavior. These methods involve: (1) mixing soil with chemical, and (2) forming columns lime in the soil without mixing. The second method is "Chemical-lime pile" or "Chemico-Pile" and widely used in Japan, and Singapore. This method is a fast, effective and well-recognized technique for ground improvement in soft clay deposit. A test embankment site in Bangkok, Thailand, improved with chemico-pile was selected for numerical modeling. The Finite Element Program PLAXIS was utilized to model the chemico-pile supported embankment. This paper presents the results from numerical modeling, and compares them with field settlement data, for both improved and natural ground.

The reinforced piled embankment (RPE) is a geotechnial composite system, consisting of five elements: piles, caps, mat with geo-grids and gravels, embankment fill and ground. This embankment system is applied for reducing lateral and vertical deformations of embankments constructed on weak foundation soils. The behavior of RPE is different from that of pile foundation or composite foundation, which has not been fully understood yet. Base on the principle of the plane of equal settlement, load transfer mechanisms between soils and piles through pile caps are investigated. This method can consider the size of caps, pile spacing, the performance of embankment fill and geogrids. The parameters, affecting the performance of embankments, such as differential settlements, cap size, and geo-grids are discussed respectively. The proposed method can conveniently obtain the ratio of loads carried by caps (piles) and the settlement of RPE.

Geosynthetic-reinforced pile-supported embankments, which contain a reinforced earth platform above the piles, are relatively new earth structures developed in the recent years. The reinforced earth platform includes one or more layers of geosynthetics at the base of the embankment. The weight of the embankment and the surcharge above are largely transferred onto the piles so that the soft soil between piles carry less load and the total and differential settlements of the embankment are reduced. A numerical study was conducted to investigate the deformation behavior of geosynethic-reinforced embankments on deep-mixed columns by considering five major influence factors: the elastic modulus of deep-mixed columns, clear spacing between columns, elastic modulus of soft soil, and tensile stiffness and layer of geosynthetics. Numerical results suggested that the inclusion of geosynthetics in the earth platform can enhance the stress transfer from soil to columns and reduce the maximum and differential settlements on the crest of the embankment, and the maximum and differential settlements on the crest of the embankment decrease with the number of geosynthetic layer and the elastic modulus of soft soil and pile material.

Dam-reservoir interaction plays an important role in the seismic response of dams. In this paper, the Scaled Boundary Finite Element Method (SBFEM) put forward by Wolf and Song is used to evaluate the dynamic water pressure acting on the dam surface during earthquakes. The basic formulas are derived. This method combines the advantages of the Finite Element Method (FEM) and the Boundary Element Method (BEM). Only the dam-water interface needs to be discretized, the computational effort is reduced significantly. In addition, the Sommerfeld radiation condition at infinity is satisfied exactly. Numerical examples of gravity dams and arch dams show that this approach is simple, effective and appropriate for engineering purpose.

This research was initiated from the Shanghai F1 speedway project constructed on the typical soft ground of Shanghai. To meet the need for its high speed, piled embankments were used to support this speedway. The embankment fill includes not only lime fly-ash but also EPS, which was used to alleviate the total load. In addition, large-size pile caps were used. All these measures were intended to reduce the ground settlement. Arching effect is one of the main reasons why the piles with individual caps can support most part of load. This paper discusses several available soil arching theories. Due to the complexity of this problem, numerical analysis was done to investigate the arching effect in embankments where the fill was uniform. The comparison between the numerical results and these available methods is presented. This study also investigates the influence of EPS on the effect of arching using the numerical method.

A three dimensional distinct element method (DEM) for numerical analysis of structural responses such as rolled concrete dam (RCD) under blast loading is presented in this paper. It is shown that the presented DEM is suitable for structural analysis by the comparison of numerical and experimental results of a RCD. The effects of some factors on the numerical results are investigated.

This paper describes an experimental study done at the Qinghai-Tibet railway. The experiment aims to reduce the temperature change inside railway embankments built in a permafrost area to minimize the embankment settlement. Riprap was placed on the slopes of the tested embankments, which is so-called air-cooled embankments. Test results show that the air-cooled embankments have better performance in stabilizing ground temperature, raising maximum thawing depth, and reducing embankment settlement as compared with ordinary embankments without riprap coverage in a high-temperature and fine-grained permafrost region. This study concludes that the air-cooled embankment, as an effective measure for protection of frozen soil, is convenient in use not only for railway construction but also for future maintenance.

Geosynthetic-reinforced and pile-supported earth platforms, as a new technique, provide an economic and effective solution for embankments, retaining walls, etc. constructed on soft soils; especially when rapid construction and/or strict deformation of the structure are required. Since it is different from traditional ones as pile-supported earth platform or geosynthetic reinforcement, a new method was developed to calculate its bearing capacity. In this method, such platforms were divided into inter-pile cells by adjacent piles and each cell has a column of soils, geosynthetics and piles. The equilibrium equations were established by examining the equilibrium of each cell and formulas for calculating its bearing capacity were obtained. The proposed method was validated by comparing the predictions with field observations and the factors influencing on the bearing capacity were analyzed.

In this study, a newly modified soil nailing technology named as the PSN (Pretension Soil Nailing), is developed to reduce both facing displacements and ground surface settlements in top-down excavation process as well as to increase the global stability. Up to now, the analytical procedure and design technique are proposed to evaluate maximum pretension force and stability of the PSN system. Also proposed are techniques to determine the required thickness of a shotcrete facing and to estimate probability of a failure against the punching shear. Based on the proposed procedure and technique, effects of the radius of a influence circle and dilatancy angle on the thickness of a shotcrete facing, bonded length and safety factors are evaluated. In addition, effects of the reduction of deformations expected by pretension of the soil nails are examined in detail throughout an illustrative example and FLAC^2D program analysis. Numerical analysis is carried out to determine a postulated failure surface as well as a minimum safety factor of the proposed PSN system using the shear strength reduction technique with the FLAC^2D program. Global minimum safety factors and local safety factors at various excavation stages computed in case of the PSN system are analyzed throughout comparisons with the results expected in case of the general soil nailing system.

Seepage pressure is important for sea wall safety monitoring. Analyzing its survey data correctly is necessary and very useful to ensure the safety of sea walls. Sea walls work in complex environment with important functions. Because of the effect of tidewater and the rain sensitivity of building materials, sea wall seepage pressure has special variation patterns that must be studied particularly. In order to comprehend the seepage rule and the relationship between it and influencing factors, rainfall amount and its duration and tidewater and its action time have been analyzed according to survey data. A former tidewater factor of seepage pressure is confirmed by statistical distance analysis, and a new rain effect factor is presented. Furthermore, in order to describe and forecast sea wall seepage pressure with high accuracy, the causal monitoring models are developed on the basis of influencing factor analysis. Shanghai Pudong sea wall seepage pressure monitoring results have been introduced. Above methods have been applied successfully in Pudong sea wall project in Shanghai, China, and the models have good precision in simulation and forecast.

Reinforced retaining walls have a ductile characteristic thus increasing the stability as the wall deforms. The deformation of the wall should be properly predicted to establish a rational design method for the reinforced retaining walls. The sliding block analysis proposed by Newmark is a useful tool for calculating seismic displacements of a slope because of its simplicity, leading to save labor of calculation. In this paper, the Newmark's method expanded to use the multi layered blocks was newly proposed and verified by the comparison with the results of centrifuge model tests conducted on reinforced retaining walls with multiple surface plates. The results show that the displacements with the expanded model of Newmark's method were good agreement with the experiments. In addition, the multi layered blocks were confirmed to show tilting deformation mode.

Based on the characteristics of earth pressure changing with the displacement of the retaining wall, a new formula for earth pressure considering wall deformation is proposed. The three parameters used in the proposed formula are calculated by the classic Rankine earth pressure coefficients or back-calculated by three sets of measured values of displacement and earth pressure. The formula is verified by the centrifugal formula test. Comparing the calculated earth pressures employing the proposed formula with those measured by the centrifugal testing, the difference is negligible, less than five percent.

The current design of mechanically stabilized earth (MSE) walls requires that the walls be "globally" stable. However, guidelines of the Federal Highway Administration and the American Association of State Highway and Transportation Officials provide only minimal guidance on global stability analyses, especially for multi-tiered MSE walls. Based on a series of analyses for a number of multi-tiered walls designed and built for the Texas Department of Transportation (TxDOT), this paper presents a procedure for a global stability analysis of multi-tiered MSE walls using a general purpose slope stability analysis program known as UTEXAS4. Analyses results also reveal the discrepancy between the design approach and analysis methodology.

This paper theoretically analyzed the shape of minor principal stress arch considering the effect of soil internal friction angle on the inclination of slip plane behind a retaining wall and the partial development of wall friction. This study shows that the shapes of minor principal stress arch behind the retaining wall are different from those proposed by others. Average vertical earth pressures were computed according to the shape of minor principal stress arch. The coefficient of earth pressure versus the angle of internal friction and the angle of wall friction was obtained. Using the coefficient in the method of horizontal differential element to compute active earth pressures on retaining wall, the theoretical formulae of the unit earth pressures, the resultant force and the points of application of the resultant force were derived. This proposed method is compared with the methods by others and some experimental data.

This paper presents observations from experiments on models of reinforced soil retaining walls subjected to horizontal base shaking. A computer controlled hydraulically driven single degree of freedom shake table is used in these tests. The test wall is constructed in a laminar box to a size of 750 mm x 500 mm in plan and 600 mm deep. The model retaining walls are constructed using wrap-around technique. Each model is instrumented with accelerometers, soil pressure sensors and LVDTs at different locations. The response of the reinforced retaining walls with the variation in the acceleration and frequency of base shaking and surcharge loading has been monitored and discussed. Increase in surcharge pressure caused attenuation in acceleration response and decrease in facing deformations. Low frequency motions tend to produce larger deformations.

Within the framework of limit equilibrium approach, the problem of limit earth pressure (active and passive) on rigid retaining walls is formulated in terms of the calculus of variations by means of Lagrange multipliers and described as the functional extreme-value problem of two undetermined function arguments. This problem is further transformed into the determination of the minimal and maximal (minimax) solutions of restrained functions incorporating the geometrical relations of the force system of the problem. The function of fmincon in the optimization toolbox of MATLAB 6.1 can be used to find the minimax solutions. Computation results show two failure modes exist, which include the sliding along a plane surface (translation) and the rotation about log-spiral cylinder surface (rotation) when the soil behind retaining walls reaches the critical state. In case of the translational mode of failure, the obtained result is the same as that of Coulomb while in case of the rotational mode of failure, the obtained result agrees well with that of Chen and Sokolovskii.

A 45m deep excavation pit with poor construction conditions is presented in this paper. Numerical methods and a monitoring system are used to control the deformation of the pit. Based on measured deformation and m-method, optimum back analysis is used to obtain the engineering properties of soils, which then are applied to predict the deformation of the retaining structure in next phrase excavation. In addition, 3D nonlinear FEM is used to simulate construction process and the deformation of the retaining structure is presented. The results of the two numerical methods match well the measured. The work of this paper provides a theoretical basis and practical reference for the design and construction of similar structures in future.