Computers and Geotechnics

Published by Elsevier
Print ISSN: 0266-352X
Inhomogeneities caused by end restraint and insufficient drainage during conventional compression triaxial tests are analysed by a numerical method. A finite element model is presented to simulate the testing procedure. The soil-platen interaction is represented by contact elements which allow frictional sliding between contacting nodes. The soil mass is represented by the modified Cam clay model. Coupled hydro-mechanical analyses are carried out in order to simulate both drained and undrained tests. The distributions of stresses and strains in the specimen for different end conditions are compared with the ideal case where no end restraint exists, in order to find representative measuring positions in the sample. Different rates of axial strain are tested in order to study the inhomogeneities caused by insufficient drainage during drained tests. Simulated results show that both end restraint and insufficient drainage can cause the barrel-shape deformation of the specimen. Stress-strain and strength properties based on global measurements are not a good representation of the true material behaviour of one single soil element at constitutive level.
Based on 2D FE analyses, simulating the excavation of a tunnel and subsequent lining with shotcrete, the impact of the employed soil model on the predicted displacements and stresses in the soil mass as well as on the predicted sectional forces in the shotcrete lining is investigated. In particular, four different soil models are considered: linear–elastic constitutive relations, the elastic–plastic models according to the Drucker–Prager and to the Mohr–Coulomb criterion as well as an elastic–plastic cap model. The computed results are compared with available field data for the vertical strains.
Considerable research works have been carried out in recent years to predict the behaviour of piled bridge abutments on soft ground. However, most of the studies are analyses based on model tests and design method proposals based on the tests in real scale are seldom performed. In this study, the results of two in situ prototype tests, where the real piled bridge abutments were constructed on soft ground and the behaviours of piles, the abutments and the ground were carefully monitored, were simulated by use of a self-coded two-dimensional FEM program incorporating the Cam-clay model and Biot's theory to take into account the coupling of elasto-plastic deformation and consolidation of the soft ground. On the basis of the results of back analyses, a simple 2D FEM-based design method for the piled bridge abutment on soft ground was proposed.
A three-dimensional model for soil-transportation structures is presented. This model exploits the geometrical periodicity of the system and takes into account the dynamic soil–structure interaction with a methodology coupling a boundary element method for the soil and a finite element formulation for the structure. A general overview of this approach is given based on several real transportation structures. Moreover, comparative studies between the different structures have been carried out. Then the model is improved by introducing a general rule for the determination of the optimal number of cells. Finally, the periodic modes propagation is investigated offering a first seizing of the significant dynamical phenomena in the soil–structure system.
The minimum safety factor and the critical slip surface of a slope can be located using various methods of optimization or random search. However these methods have not been adopted to search for the three-dimensional (3D) critical slip surface. This paper proposes a new Monte Carlo random simulation method to identify the 3D critical slip surface, in which assuming the initial slip as the lower part of an ellipsoid, the 3D critical slip surface in the 3D slope stability analysis is located by means of a minimization of the 3D safety factor. Based on a column-based three-dimensional slope stability analysis model, a new geographic information systems (GIS) grid-based 3D deterministic model has been developed to calculate the 3D safety factor. Several practical cases, of known minimum safety factor and its critical slip surface by using a two-dimensional (2D) optimization or random technique, have been extended to 3D slope problems to locate the 3D critical slip surface and to compare with the 2D results. As another 3D case, a multi-strata slope, considering the effect of a fault and the underground water, is used to demonstrate the efficiency and capability of the proposed method. Compared with the 2D result, the resulting 3D critical slip surface has no apparent difference for simple slope geometries, but the associated 3D safety factor is definitely higher.
The aim of this paper is to evaluate the influence of negative skin friction in pile foundations. Three dimensional nonlinear analyses for a single pile and pile groups were carried out for a specific case and some case studies as well. Contrary to simplified conventional analysis in which the predictions are usually overestimated and could be considered as an upper limit, it was found that the dragload of a pile in a group depends on the surface load, the pile configuration, the pile position in a group, the ultimate skin friction and the interface stiffness. It has been demonstrated that for fixed-head friction pile groups the dragload group effect is significantly greater than in the case of free-head end-bearing pile groups. Moreover, predictions for internal piles have shown considerably smaller dragloads for fixed-head piles, which is in accordance with experimental findings. It has also been demonstrated that when the construction of an embankment precedes the application of the foundation working load, the effect of negative skin friction is considerably smaller than in the reverse case.
Full-height piled bridge abutments constructed on soft clay are prone to soil–structure interaction effects. A series of geotechnical centrifuge tests of this type of structure has been undertaken, and an accompanying series of plane strain finite element analyses are reported. Some aspects of the structure do not conform to a plane strain analysis (most notably the piles), and the methods used to incorporate this soil–structure interaction are described. Success of the methods is illustrated by good comparison with the centrifuge test results, and the numerical analyses revealed interaction effects which could not be specifically identified in the centrifuge tests.
In finite element analysis of soil-structure interaction problems involving firm to stiff overconsolidated clay, there have been difficulties in modelling the stress-strain response of the soil. Non-linearity and anisotropy of the soil depend on the inherent anisotropy of its particle structure and the induced anisotropy of its stress history and current stress path. In CRISP modelling of the centrifuge test of an abutment wall and its backfill of sand on the surface of a firm to stiff overconsolidated kaolin, the clay foundation was divided into 6 broad zones in accordance with the stress history and stress path. Undrained movements of the abutment and its subsoil were closely modelled in two analyses; one with a non-linear elastic model and the other with the Schofield model with shear modulus G assigned to the foundation zone in accordance with the estimated strain level as well as stress history and stress path. In the prediction of consolidation movement, there is a difficulty in the current critical state soil model in CRISP. The fe solution incorrectly predicted that substantial horizontal movement would accompany settlement due to consolidation, whereas the centrifuge test showed mainly vertical movement. This is attributable to the pronounced anisotropy separately observed in element tests.
A numerical method for liquefaction analysis of saturated soils with large deformation is presented. Formulations are based on Biot's two-phase mixture theory and the updated Lagrangian method. Governing equations consist of an equilibrium equation and one for mass conservation. Based on the u-p formulation, displacement of the soil skeleton and pore pressure are the two basic unknown variables. Governing equations are discretized by the FE-FD coupled method. A cyclic elasto-plastic constitutive model is used to describe the liquefaction behavior of saturated soils under dynamic loading. The coefficient of porosity is considered to vary with large deformation. The coefficient of permeability is assumed to be a function of the void ratio. The examples show the flexibility and applicability of the proposed method. Comparison is made between large and small strain solutions. The large deformation analysis shows that liquefaction occurs earlier in seabed deposits under wave action and deeper in the soil surrounding an embankment subjected to strong earthquake motion than it does in small strain analysis.
A high-cycle explicit model for the accumulation of strain in sand due to small cyclic loading is presented. The dependence of the accumulation rate on stress, void ratio, cyclic history and the type of loading is discussed. In particular, the ovality and the polarization of the strain path during a cycle are considered. Attention is given to the theoretical aspects of the constitutive description of the cumulative settlement and to the FE-implementation. The essential experimental results are also presented. Finally, an example of an FE-analysis of a strip foundation under cyclic vertical loading is given.
Wave attenuation across fractured rock masses is a great concern of rock engineers to assess the safety of underground structures in and on rocks under dynamic loads. Due to the discreteness of rock masses, the universal distinct element code (UDEC) has been adopted for the study of rock mass problems. In this paper, the calibration work of UDEC modelling on P-wave propagation across single linearly and nonlinearly deformable fractures is conducted. Subsequently, numerical studies of P-wave propagation across multiple nonlinearly deformable fractures are carried out. The magnitude of transmission coefficient is calculated as a function of nondimensional fracture spacing for different numbers of fractures. The results reveals that under some circumstances, the magnitude of transmission coefficient not only increases with increasing number of fractures, but also is larger than 1.
This paper describes a two-dimensional finite element model developed to simulate the wave-induced hydrodynamic uplift force acting on a submarine pipeline buried in sandy seabed sediments subjected to continuous loading of sinusoidal surface waves. Neglecting inertia forces, a linear-elastic stress-strain relationship for the soil, and Darcy's law for the flow of pore fluid are assumed. The model takes into account the compressibility of both components (i.e. pore fluid and soil skeleton) of the two-phase medium. The governing equations are discretized using the Galerkin finite element method. Due to the geometry of the problem, four-node isoparametric elements are chosen and the Gaussian quadrature formulae is used in a numerical integration procedure to compute the element stiffness matrices. Several verification problems are presented to demonstrate the model utility and check numerical accuracy influenced by the time- and space-discretization as well as the quadrature rule of numerical integration.
Error estimation and h-adaptive finite element procedures are implemented for large deformation analyses of foundations on soil where the strength increases with depth. Errors are estimated by comparing strains at Gauss points with more accurate estimates using Superconvergent Patch Recovery (SPR). Mesh refinement using subdivision concept is then used iteratively to obtain optimal meshes, satisfying a minimum element size criterion. The Remeshing and Interpolation Technique with Small Strain (RITSS) approach is then used for large deformation analysis, with stress interpolation using either a modified form of the unique element method (MUEM), or the stress-SPR approach. Example analyses are then presented illustrating the three main aspects of mesh refinement, stress interpolation and large deformation response. Criteria are given for minimum element size and displacement increment for strip and circular foundations bearing on soil with varying degree of non-homogeneity, and computed bearing capacities are shown to compare well with lower bound estimates. The effect of soil weight on deep penetration of a strip foundation is discussed, with particular reference to the pattern of soil heave adjacent to the foundation, and the magnitude of the bearing capacity.
The concept of using neural networks in constitutive modeling has been proposed by the first author and his co-workers. In this methodology, neural networks are trained directly with the results of material tests, and the trained neural networks can be used in analysis of boundary value problems similar to any other material model. In this paper, we introduce nested adaptive neural networks, a new type of neural network developed by Ghaboussi and his co-workers, and apply this neural network in modeling of the constitutive behavior of geomaterials. Nested adaptive neural networks take advantage of the nested structure of the material test data, and reflect it in the architecture of the neural network. This new neural network is applied in modeling of the drained and undrained behavior of sand in triaxial tests.
An attempt has been made to investigate the possibility of using adaptive network-based fuzzy inference systems to predict the post-construction settlement of rockfill dams. Four types of dams, namely, central core, sloping core, compacted membrane faced, and dumped membrane faced rockfill dams are considered in this study. An index is defined to indicate the combined compressibility of the dam embankment and its foundation material. Therefore, three variables representing dam height, dam type, and dam compressibility are used as input variables for the network. A database, prepared from reported post-construction settlement of rockfill dams, has been used to train the neurofuzzy inference system and to determine the network parameters. Performance of the trained network is compared with the conventional methods. The comparison indicates that the proposed modelling is more reliable and has a better performance than conventional methods.
A technique is developed for analysing elasto-plastic unbounded media by adaptively coupling the finite-element method with the scaled boundary finite-element method. The analysis begins with a finite-element mesh that tightly encloses the load–medium interface, capturing non-linearity in the very near field. The remainder of the problem is modelled accurately and efficiently using the semi-analytical scaled boundary finite-element method. Load increments are applied in the usual (finite-element) way and the plastic stress field grows outwards from the load–medium interface as the solution advances. If plasticity is detected at a Gauss point in the outer band of finite-elements, an additional band of finite-elements are added around the perimeter of the existing mesh and the scaled boundary finite-element domain is stepped out accordingly. This technique exploits the most attractive features of both the finite-element and scaled boundary finite-element methods. The technique is shown to be highly accurate and both user and computationally efficient.
This paper aims to investigate the mechanical behaviour of a hybrid reinforced earth embankment built in limited width adjacent to a slope. This embankment system incorporates reinforced earth embankments with soil nails, installed in the existing ground. Soil nails work to provide additional resisting forces to stabilize the embankment which may be unstable due to insufficient reinforcement length. Nail forces developed in the hybrid reinforced earth embankment with various geometric conditions in the fill space are analyzed. The FE method is used to simulate the construction of the hybrid reinforced earth embankment. Influence of reinforcement length, reinforcement stiffness, and slope gradient on the nail forces developed following the construction is analyzed and discussed. Additionally, design concerns for the hybrid reinforced earth embankment system are also studied. Simple charts for estimating the maximum nail force mobilized at back of the hybrid reinforced earth embankment are established in this research and can be helpful in the design of the soil nails in the system.
The short-term behaviour of pile groups subjected to lateral pressures by deformation of a clay layer under an adjacent surcharge load was studied using three dimensional finite element analysis. The main aim of the analysis was to investigate the pile-clay interaction behaviour. A load-path-dependent, non-linear constitutive model was used to describe the clay, which required knowledge of in situ stresses and recent strain history. Numerical results compared well with those from a centrifuge model test. The effects of the different in situ stresses and strains likely in prototypes and centrifuge model tests were also studied with particular interest in the load-transfer relationships and soil deformation behaviour around the piles.
In many instances, a soil may be demonstrably polluted by a chemical but not pose an immediate threat to the environment. In this case, it may be acceptable to put in place a decontamination strategy that ensures the pollutant will remain within a prescribed region and gradually be sorbed onto a substrate over a period of time. One practical way of implementing such a decontamination strategy is to simply excavate a trench around the polluted area and backfill the trench with a sorbant material. Over time, the contaminant will be transported through the soil and sorbed onto the material contained within the trench. When decontamination of the polluted soil is deemed complete, the sorbant material may be removed from the trench in readiness for ultimate disposal. The engineer responsible for the design of such a decontamination strategy would be interested in such questions as the required thickness of sorbant material within the trench, the influence of the partitioning coefficient of the sorbant material on the trench thickness, the time required for contaminant transport through the sorbant material, and the time at which decontamination of the polluted soil is deemed to be complete. This Paper employs a one dimensional dispersion-advection equation to generate non-dimensional design charts that can assist the design engineer in estimating the performance of the proposed decontamination strategy. The potential application of the design charts is demonstrated by means of illustrative examples.
This paper describes an identification process for determination of material parameters in a constitutive relationship, describing time dependency of air permeability of shotcrete tunnel lining. A numerical model has been developed to predict the air losses from tunnel face and perimeter walls in compressed air tunnelling. Field data from a Tunnel in Germany has been used to verify and calibrate the numerical model. A relationship has been established to describe the variation of the air permeability of shotcrete tunnel lining with time and the technique of parameter identification has been used to determine the parameters of this relationship. A genetic algorithm has been used in the optimisation procedure. It has been shown that time dependency of permeability of shotcrete plays a key role in controlling the air losses in driving tunnels under compressed air with shotcrete as a temporary or permanent lining and this time dependency should be taken into account in design.
This paper presents a numerical formulation for frictional contact problems associated with pile penetration. The frictional contact at the soil–pile interface is formulated using the theory of hardening/softening plasticity, so that advanced models for the interface can be dealt with. A smooth discretisation of the pile surface is proposed using Bézier polynomials. An automatic load stepping scheme is proposed, which features an error control algorithm and automatic subincrementation of the load increments. The numerical algorithms are then used to analyse the installation process of pushed-in axial piles. It is shown that the smooth discretisation of the pile surface is effective in reducing the oscillation in the predicted pile resistances and the automatic load stepping scheme outperforms the classical Newton–Raphson scheme for this type of problem.
This paper presents an overview of constitutive modelling of unsaturated soils and the numerical algorithms for solving the associated boundary value problems. It first discusses alternative stress and strain variables that can be used in constitutive models for unsaturated soils. The paper then discusses the key issues in unsaturated soil modelling and how these issues can be incorporated into an existing model for saturated soils. These key issues include (1) volumetric behaviour associated with saturation or suction changes; (2) strength behaviour associated with saturation and suction changes, and (3) hydraulic behaviour associated with saturation or suction changes. The paper also shows how hysteresis in soil–water characteristics can be incorporated into the elasto-plastic framework, leading to coupled hydro-mechanical models. Finally, the paper demonstrates the derivation of the incremental stress–strain relations for unsaturated soils and discusses briefly the new challenges in implementing these relations into the finite element method.
Cone penetration test (CPT) is one of the most common in situ tests which is used for pile design because it can be realized as a model pile. The measured cone resistance (qc) and sleeve friction (fs) usually are employed for estimation of pile unit toe and shaft resistances, respectively. Thirty three pile case histories have been compiled including static loading tests performed in uplift, or in push with separation of shaft and toe resistances at sites which comprise CPT or CPTu sounding. Group method of data handling (GMDH) type neural networks optimized using genetic algorithms (GAs) are used to model the effects of effective cone point resistance (qE) and cone sleeve friction (fs) as input parameters on pile unit shaft resistance, applying some experimentally obtained training and test data. Sensitivity analysis of the obtained model has been carried out to study the influence of input parameters on model output. Some graphs have been derived from sensitivity analysis to estimate pile unit shaft resistance based on qE and fs. The performance of the proposed method has been compared with the other CPT and CPTu direct methods and referenced to measured piles shaft capacity. The results demonstrate that appreciable improvement in prediction of pile shaft capacity has been achieved.
In application to numerical analysis of geotechnical problems, the limit-state surface is usually not known in any closed form. The probability of failure can be assessed via the so-called reliability index. A minimization problem can naturally be formed with an implicit equality constraint defined as the limit-state function and optimization methods can be used for such problems. In this paper, a genetic algorithm is proposed and incorporated into a displacement finite element method to find the Hasofer–Lind reliability index. The probabilistic finite element method is then used to analyse the reliability of classical geotechnical systems. The performance of the genetic algorithm (GA) is compared with simpler probability methods such as the first-order-second-moment Taylor series method. The comparison shows that the GA can produce the results fairly quickly and is applicable to evaluation of the failure performance of geotechnical problems involving a large number of decision variables.
Extrapolation method: stress increments  
The numerical integration of the stress–strain relationship is an important part of many finite element code used in geotechnical engineering. The integration of elasto-plastic models for unsaturated soils poses additional challenges associated to the presence of suction as an extra constitutive variable with respect to traditional saturated soil models. In this contribution, a range of explicit stress integration schemes are derived with specific reference to the Barcelona Basic Model (BBM), which is one of the best known elasto-plastic constitutive models for unsaturated soils. These schemes, however, do not address possible non-convexity of the loading collapse (LC) curve and neglect yielding on the suction increase (SI) line. The paper describes eight Runge–Kutta methods of various orders with adaptive substepping as well as a novel integration scheme based on Richardson extrapolation. The algorithms presented also incorporate two alternative error control methods to ensure accuracy of the numerical integration. Extensive validation and comparison of different schemes are presented in a companion paper. Although the algorithms presented were coded for the Barcelona Basic Model, they can be easily adapted to other unsaturated elasto-plastic models formulated in terms of two independent stress variables such as net stress and suction.
The concept of response surface method (RSM) is used to generate approximate polynomial functions for ultimate bearing capacity and settlement of a shallow foundation resting on a cohesive frictional soil for a range of expected variation of input soil parameters. The response surface models are developed using available conventional equations and numerical analysis. Considering the variations in the input soil parameters, reliability analysis is performed using these response surface models to obtain an acceptable value of the allowable bearing pressure. The results of the reliability analysis are compared with the results of Monte Carlo simulation and it is demonstrated that application of response surface method in the probabilistic analysis can considerably reduce the computational efforts and memory requirements. It is also concluded that conventional analysis using available equations and numerical analysis when used in conjunction with reliability analysis enable a rational choice of allowable pressure and help in decision-making process.
The main purpose of the paper is the numerical analysis of seismic site effects in Caracas (Venezuela). The analysis is performed considering the boundary element method in the frequency domain. A numerical model including a part of the local topography is considered, it involves a deep alluvial deposit on an elastic bedrock. The amplification of seismic motion (SH-waves, weak motion) is analyzed in terms of level, occurring frequency and location. In this specific site of Caracas, the amplification factor is found to reach a maximum value of 25. Site effects occur in the thickest part of the basin for low frequencies (below 1.0 Hz) and in two intermediate thinner areas for frequencies above 1.0 Hz. The influence of both incidence and shear wave velocities is also investigated. A comparison with microtremor recordings is presented afterwards. The results of both numerical and experimental approaches are in good agreement in terms of fundamental frequencies in the deepest part of the basin. The boundary element method appears to be a reliable and efficient approach for the analysis of seismic site effects.
Artificial neural networks are capable of learning complex nonlinear relationships from a large amount of accumulated data, and similar to human brains, are noise and fault tolerant. This unique capacity suggests that neural networks would be very useful in certain geotechnical engineering applications. A back-propagation network is set up and trained to predict the pile bearing capacity from dynamic testing data. The trained network produces better results than a pile driving formula approach. The effects of various network parameters on the network results are examined in detail. The general understanding developed is potentially useful for the application of neural networks in other geotechnical engineering problems.
Fully explicit analytical solutions are developed for one-dimensional large strain consolidation in both thick and thin soil layers. Numerical examples are given and comparisons are made with the classical small strain theory. It is shown that, unlike the results in classical small strain theory, the average degree of consolidation defined by stress (i.e. Up) and that defined by strain (i.e. Us) in large strain theory are different. The magnitude of settlement predicted by large strain theory is found to be smaller while both the development of settlement and the dissipation of excess pore water pressure (as shown by Us and Up) are found to be faster than in small strain consolidation in the cases studied. Another interesting observation is that the discrepancy between large and small strain theories diminishes with reducing compressibility (i.e. increasing stiffness) of soil and decreasing magnitude of applied load.
A semi-analytical and semi-numerical method is developed for the analysis of plate-layered soil systems. Applying a Hankel transform, an expression relating the surface settlement and the reaction of the layered soil is derived. Such a reaction can be treated as a load acting on the plate in addition to the applied external load. Having the plate modeled by eight-noded isoparametric elements, the governing equations of the plate can be formed and solved. Numerical examples, including square, trapezoidal and circular plates resting on elastic layered soil, are given to demonstrate the advantages, accuracy and versatility of this method.
Assessment of the stability of embankment and cut slopes over the life of a project are critical issues for railway and motorway infrastructure projects. Experience has shown that many slope failures occur during or shortly after rainfall. Analyses show that failure is initiated by the reduction in near surface suction over some critical depth in the slope. A simple method is proposed in this paper to estimate the time needed for a wetting front to develop. The method which is a modification of the traditional Green–Ampt infiltration model assumes that ponding of water cannot occur on soil slopes and as a consequence soil in the wetted zone remains partially saturated at the point of slope failure. It differentiates between cases where the initial suction in the slope is high and the rate of infiltration is controlled by the rainfall intensity (supply controlled) and, situations where the suction is low, and the rate of infiltration is controlled by the infiltration capacity of the soil (demand controlled). When applied to a case history where field measurements of infiltration into a slope were available the new method provided a reasonable approximation of the measured infiltration time.
The behavior of vertical anchor plates embedded in reinforced and non-reinforced cohesionless soil has been investigated with the help of small-scale model tests. Steel rods (model piles) with different lengths and diameters placed vertically or inclined at different locations relative to the anchor plate were used to reinforce the sand in front of both strip and square anchor plates. The considered parameters include pile length, pile diameter, pile spacing, position of pile row relative to the anchor plate and the inclination angle of the installed piles, the anchor embedment depth, the anchorage geometry and the relative density of sand. The test results indicate that this type of reinforcement significantly increases the stiffness of the soil and the pullout resistance of shallow anchor plates. Based on test results, critical values were discussed and recommended.
Following a brief review of the physico-mechanical properties of soils, this work analyzes and comments upon some of the most frequently used approaches in anchored sheet pile wall design. The analysis highlights the conceptual differences between the various approaches, often leading, inevitably, to markedly diverging results. Although the probabilistic approach cannot be applied extensively, mainly because it is difficult to obtain statistical modelling of the soil mass, it nevertheless enables designers to avoid certain ambiguities that are present in the commonly used approach based on the Safety Factor. In addition, the probabilistic approach also permits handling of the calibrations required for the approaches to partial coefficients to be effective and applicable to different local conditions associated with the diversity of soils, different modes of construction, etc. Numerical results obtained for a simple probabilistic model lead to conclusions which are certainly not exhaustive but may contribute significant elements for reflection.
The construction of sheet pile walls may involve either excavation of soils in front or backfilling of soils behind the wall. These construction procedures generate different loading conditions in the soil and therefore different wall behavior should also be expected. The conventional methods, which are based on limit equilibrium approach, commonly used in the design of anchored sheet pile walls do not consider the method of construction. However, continuum mechanics numerical methods, such as finite element method, make it possible to incorporate the construction method during the analyses and design of sheet pile walls. The effect of wall construction type for varying soil conditions and wall heights were investigated using finite element modeling and analysis. The influence of construction method on soil behavior, wall deformations, wall bending moments, and anchor forces were investigated. The study results indicate that walls constructed by backfill method yield significantly higher bending moments and wall deformations. This paper presents the results of the numerical parametric study performed and comparative analyses of the anchored sheet pile walls constructed by different construction methods.
Marquees and other temporary light structures are connected to the ground by tensile anchors that resist uplift forces. The existing methods for predicting the pullout capacity of these anchors are inaccurate and incomplete. As a result, failures of such structures are not rare and have resulted in deaths and tens of thousands of dollars of damage. This paper aims to increase the safety of temporary light structures, such as marquees, by developing a more accurate pullout capacity prediction method based on artificial neural networks (ANNs). Two types of ANNs are examined, namely, multi-layer perceptrons (MLPs) that are trained with the back-propagation algorithm and B-spline neurofuzzy networks that are trained with the adaptive spline modeling of observation data (ASMOD) algorithm. In order to facilitate the use of the MLP model, it is made available in a tractable equation form. Predictions of pullout capacity from the developed ANN models are obtained and compared with values predicted by traditional methods currently used in practice. The results indicate that ANNs are able to predict accurately the pullout capacity of ground anchors and outperform the existing methods.
The vertical uplift resistance of two interfering rigid rough strip anchors embedded horizontally in sand at shallow depths has been examined. The analysis is performed by using an upper bound theorem of limit analysis in combination with finite elements and linear programming. It is specified that both the anchors are loaded to failure simultaneously at the same magnitude of the failure load. For different clear spacing (S) between the anchors, the magnitude of the efficiency factor (ξγ) is determined. On account of interference, the magnitude of ξγ is found to reduce continuously with a decrease in the spacing between the anchors. The results from the numerical analysis were found to compare reasonably well with the available theoretical data from the literature.
Hypoplasticity has been proven to be successful in reproducing the mechanical behaviour of sands. However, in case of clays the incremental stiffness in shear is predicted too low. An investigation of the hypoplastic equation based on the analysis of response envelopes reveals the reason of this deficiency. An improvement is proposed and compared with experimental results for Rio de Janeiro clay.
The Sekiguchi–Ohta model is extended to be a unified three-dimensional elastoplastic model for clays, silts and sands by introducing a transformed stress tensor and a new hardening parameter H. The model can describe the negative and positive dilatancy of soils with an initially stress-induced anisotropy in three-dimensional (3D) stress. An elastoplastic constitutive tensor is derived for the application of the proposed model to finite element (FE) analyses. A FE analysis example for a test embankment is given and the result demonstrates that the proposed model can predict reasonably the deformation of anisotropically consolidated clay and sand layers under embankment loading.
In this paper, we propose an anisotropic plastic damage model for semi-brittle geomaterials based on a discrete thermodynamic approach. The macroscopic plastic deformation is generated by frictional sliding of weakness planes. The evolution of damage is related to growth of such weakness planes. The local frictional sliding in each family of weakness planes is described by a non-associated plastic model taking into account material softening and volumetric dilatancy. The damage evolution is coupled with plastic deformation and modelled by an isotropic damage criterion. The proposed model is applied to modelling mechanical responses of typical sandstone under different loading paths. There is good agreement between numerical predictions and experimental data. Further, the anisotropic distributions of plastic deformation and induced damage are analysed and discussed.
The contribution of fibers to the strength of fiber-reinforced soils is very much dependent on the distribution of orientation of the fibers. The fibers in the direction of largest extension contribute most to the strength of the composite, whereas the fibers under compression have an adverse effect on the composite stiffness, and they do not produce an increase in the composite strength. Considering a contribution of a single fiber to the work dissipation during failure of the composite, and integrating this dissipation over all fibers in a composite element, a failure criterion is derived for fiber-reinforced sand with an anisotropic distribution of fiber orientation. A deformation-induced anisotropy was detected in experiments. Specimens with initially isotropic distribution of fiber orientation exhibited a kinematic hardening effect. The evolution of fiber orientation in the deformation process was found to have been the cause of the anisotropic hardening.
Experiments indicate that in one-dimensionally consolidated natural clays the elastic anisotropy is much stronger than the plastic strain anisotropy. Moreover, the elastic anisotropy appears to be dependent on the pre-consolidation strain. Coupled elasto-plastic constitutive law is shown to be able to simulate these anisotropy effects of natural clay deposits. In this law the elastic potential is not only a function of stress, but additionly of the plastic strain. The plastic strain comprises the geological process of pre-consolidation idealized as an one-dimensional plastic straining as well as a mechanically induced strain due to engineering activity. Calibration of the model and simulation of some stress paths are presented and related to the classical experimental results by Mitchell (1972).
The possibilities and limitations of modelling the behaviour of soft natural clays within the multi-laminate framework are discussed in this paper. The presented version of the constitutive model was developed to describe two important characteristics of the soft clay behaviour: strength anisotropy and destructuration. Viscous effects are not taken into account. Strength anisotropy of the soft clay is achieved using the multi-laminate model with controlled spatial distribution of the preconsolidation pressure over the sampling planes. This spatial distribution is related to the initial stress state and is additionally adjusted by the distribution of bonding. Bonding or structural anisotropy is not permanent in the model and may be reduced during the process of destructuration. Interrelation between stress-induced and bonding anisotropy is shown by investigating the shape changes of a macro yield surface resulting from the multi-laminate modelling. Performance of the new model is demonstrated by standard element tests and also by the boundary value problem of an embankment on soft ground.
Naturally occurring cohesive soil deposits are inherently anisotropic with respect to different properties amongst which is the shear strength. This anisotropy is primarily due to the process of sedimentation followed by predominantly one-dimensional consolidation. The effect of strength anisotropy on the stability of slopes has commonly been investigated using the mass procedure in which the mass of soil failed is taken as a one unit. The analysis is conducted by utilizing either the conventional limit equilibrium method or the upper bound technique of limit plasticity. It is, however, the method of slices that is most widely employed in the analysis of slope stability problems, especially with the advent of the digital computer. In this study, the influence of cohesion anisotropy on the stability of slopes in homogeneous soil, with internal friction angle greater than zero, was investigated using the method of slices. Analysis was made for the case of two-dimensional circular slip surfaces. It has been found that soil anisotropy has a significant effect on the stability of slopes when the slope angle is less than 53°. However, such an effect becomes insignificant as the soil angle of internal friction exceeds about 10°. The results also indicated that the geometry of the slip surface is only slightly affected by the degree of anisotropy in the soil cohesion.
The stress–strain and volume change behavior of sand and gravel under drained triaxial compression test conditions was modeled using feed-back artificial neural networks. A large experimental database obtained from published literature was used in training, testing, and prediction phases of three neural network based soil models. Issues related to the number of hidden units, magnitude of strain increment during feed-back, and over-training error are discussed. These models can accurately represent the effects of mineralogy, grain shape and size distribution, void ratio, and confining pressure. The observed behavior in terms of a non-linear stress–strain relation, compressive volume change at low stress levels, and volume expansion at high stress levels are captured well by these models.
An elastic interaction analysis of annular slab with soil has been made in which the soil has been modelled as either (i) homogeneous and isotropic half-space or (ii) multilayered half-space. The interaction problem has been solved by using a combined approach (hybrid method) in which the finite element method (FEM) has been used to obtain the stiffness matrix for the annular slab while for the elastic half-space soil models, the stiffness matrix has been obtained using an analytical solution. In the analysis, the interface between the annular slab and the elastic half-space has been assumed to be frictionless. Later the combined stiffness matrix of the annular slab-elastic half-space has been obtained using the displacement continuity condition at the interface. Results of the contact pressure distribution, displacement (settlement) profile and bending moment variation in an annular slab have been presented for different half-space soil models. It is shown that the hybrid method presented here is easy to apply and computationally efficient.
The paper considers two techniques to model the Cone Penetration Test (CPT) end resistance, qc in a dense sand deposit using commercial finite element programmes. In the first approach, Plaxis was used to perform spherical cavity expansion analyses at multiple depths. Two soil models, namely; the Mohr–Coulomb (MC) and Hardening Soil (HS) models were utilized. When calibrated using simple laboratory element tests, the HS model was found to provide good estimates of qc. However, at shallow depths, where the over-consolidation ratio of the sand was highest, the relatively large horizontal stresses developed prevented the full development of the failure zone resulting in under-estimation of the qc value. The second approach involved direct simulation of cone penetration using a large-strain analysis implemented in Abaqus/Explicit. The Arbitrary Lagrangian Eulerian (ALE) technique was used to prevent excessive mesh deformation. Although the Druker–Prager soil model used was not as sophisticated as the HS model, excellent agreement was achieved between the predicted and measured qc profiles.
The problem of automatic identification of three dimensional rock blocks formed by discontinuities such as joints and faults is studied in this paper. An algorithm and its simulation results for identification of rock blocks are presented. A dynamic link list associated with the algorithm is employed to realize the representation of a polyhedron as well as its forming process by topological identification. Attention is focused on the simplification and the realization of the blocky structure recognition system. Two typical examples are given, one for a blocky structure with all convex polyhedra, and the other with polyhedra of convex and concave geometries. The rock block generator is served as a pre-processor for discrete element method (DEM) using polyhedral elements or for other methods such as block theory (BT) and discontinuous deformation analysis (DDA). The computer simulation can also be extended to rock block identification involving curved joints and faults.
Both empirical and theoretical (thermodynamic) investigations indicate that the soil–water characteristic curve (SWCC) is an essential element in unsaturated soil modeling and it is hysteretic in nature, owing to the energy dissipation during moving fluids in or out of the soil. Because of its fundamental significance, this hysteretic feature can no longer be ignored in unsaturated soil modeling. This note presents a simple phenomenological approach for modeling hysteretic SWCCs following arbitary wetting/drying paths.
Top-cited authors
Scott William Sloan
  • University of Newcastle
Daichao Sheng
  • University of Technology Sydney
ym Cheng
  • The Hong Kong Polytechnic University
Mingjing Jiang
  • Suzhou University of Science and Technology
Chuangbing Zhou
  • Nanchang University