International Journal for Numerical and Analytical Methods in Geomechanics

A new macroscopic plasticity model for non-cohesive granular materials, with the focus on coarse-sized materials (railway ballast), is presented. The corresponding incremental relations are reformulated in the space of stress invariants that is extended with the internal (hardening) variables and the corresponding dissipative stresses. The model is calibrated using data from Conventional Triaxial Compression (CTC) tests. A function evaluation method is used for the optimization. A “multi-vector” strategy for choosing the appropriate start vector is proposed.

In this paper, the mechanism of fault pressurization in rapid slip events is analyzed on the basis of a complete characterization of the thermo-poro-mechanical behavior of a clayey gouge extracted at 760m depth in Aigion fault in the active seismic zone of the Gulf of Corinth, Greece. It is shown that the thermally collapsible character of this clayey gouge can be responsible for a dramatic reduction of effective stress and a full fluidization of the material. The thickness of the 'ultra localized' zone of highly strained material is a key parameter that controls the competing phenomena of pore pressure increase leading to fluidization of the fault gouge and temperature increase leading to pore fluid vaporization.

A micro-hydromechanical model for granular materials is presented. It combines the discrete element method (DEM) for the modeling of the solid phase and a pore-scale finite volume (PFV) formulation for the flow of an incompressible pore fluid. The coupling equations are derived and contrasted against the equations of conventional poroelasticity. An analogy is found between the DEM-PFV coupling and Biot's theory in the limit case of incompressible phases. The simulation of an oedometer test validates the coupling scheme and demonstrates the ability of the model to capture strong poromechanical effects. A detailed analysis of microscale strain and stress confirms the analogy with poroelasticity. An immersed deposition problem is finally simulated and shows the potential of the method to handle phase transitions.

An analytical solution to 1D coupled water infiltration and deformation is derived using a Fourier integral transform. Exponential functional forms are used to represent the hydraulic conductivity–pore-water pressure relationship and the soil-water characteristic curve. Fredlund's incremental-linear constitutive model for unsaturated soils is adopted. The analytical solution considers arbitrary initial pore-water pressure distributions and flux and pressure boundary conditions. The corresponding analytical solutions to coupled steady-state problems are also obtained. The analytical solutions demonstrate that the coupling of seepage and deformation plays an important role in water infiltration in unsaturated soils. In the early stages of infiltration, the difference between uncoupled and coupled conditions becomes marked over time, and in late stages, the difference caused by the coupling effects diminishes toward the steady state. The difference between the uncoupled and coupled conditions increases with decreasing desaturation coefficient (). Pore-water pressure or deformation changes caused by the coupling effects are mainly controlled by the degree of soil volume change due to a change in soil suction (H). The smaller the absolute value of H, the greater the effect of coupling on the infiltration and deformation. The ratio of rainfall intensity to saturated permeability (q/ks) also has a strong influence on the coupled seepage and deformation. Copyright © 2008 John Wiley & Sons, Ltd.

An analytical solution to 1D coupled water infiltration and deformation in layered soils is derived using a Laplace transformation. Coupling between seepage and deformation, and initial conditions defined by arbitrary continuous pore-water pressure distributions are considered. The analytical solutions describe the transient pore-water pressure distributions during 1D, vertical infiltration toward the water table through two-layer unsaturated soils. The nonlinear coupled formulations are first linearized and transformed into a form that is solvable using a Laplace transformation. The solutions provide a reliable means of comparing the accuracy of various numerical methods. Parameters considered in the coupled analysis include the saturated permeability (ks), desaturation coefficient (α), and saturated volumetric water content (θs) of each soil layer, and antecedent and subsequent rainfall infiltration rates. The analytical solution demonstrates that the coupling of seepage and deformation plays an important role in water infiltration in layered unsaturated soils. A smaller value of α or a smaller absolute value of the elastic modulus of the soil with respect to a change in soil suction (H) for layered unsaturated soils means more marked coupling effect. A smaller absolute value of H of the upper layer soil also tends to cause more marked coupling effect. A large difference between the saturated coefficients of permeability for the top and bottom soil layers leads to reduced rainfall infiltration into the deep soil layer. The initial conditions also play a significant role in the pore-water pressure redistribution and coupling effect. Copyright © 2011 John Wiley & Sons, Ltd.

A mathematical model is developed for the dynamic analysis of earthquake-triggered rapid landslides, considering two mechanically coupled systems: (a) the accelerating deformable body of the slide and (b) the rapidly deforming shear band at the base of the slide. The main body of the slide is considered as a one-phase mixture of Newtonian incompressible fluids and Coulomb solids sliding on a plane of variable inclination. The evolution of the landslide is modeled via a depth-integrated model of the Savage–Hutter type coupled with: (a) a cyclic hysteretic constitutive model of the Bouc–Wen type and (b) Voellmy's rheology for the deformation of the material within the shear band. The original shallow-water equations that govern the landslide motion are appropriately reformulated to account for inertial forces due to seismic loading, and to allow for a smooth transition between the active and the passive state. The capability of the developed model is tested against the Higashi–Takezawa landslide. Triggered by the 2004 Niigata-ken Chuetsu earthquake, the slide produced about 100m displacement of a large wedge from an originally rather mild slope. The mechanism of material softening inside the shear band responsible for the surprisingly large run-out of the landslide is described by a set of equations for grain crushing-induced pore-water pressures. The back-analysis reveals interesting patterns on the flow dynamics, and the numerical results compare well with field observations. It is shown that the mechanism of material softening is a crucial factor for the initiation and evolution of the landslide, while viscoplastic frictional resistance is a key requirement for successfully reproducing the field data. Copyright © 2009 John Wiley & Sons, Ltd.

This paper describes the essential features of a numerical technique for the simulation of the coupled fluid flow and deformation in a 2D assembly of poroelastic blocks and transmissive fractures. The boundary element method (BEM) is applied to each block to reduce Navier and diffusion equations to a set of integral equations involving block boundary terms, whereas a Galerkin weighted-residuals finite element method (FEM) is applied to the fracture diffusion equations. In addition, fracture local equilibrium is rendered through spring-like equations relating the stresses to the relative displacements of the fracture walls. A time-marching process is implemented leading to an algebraic system where the right-hand side vector is built based on the collected solutions of the previous time steps. The technique requires the meshing of the fracture network only. The accuracy of the results is adequate even with relatively coarse meshes without the resort to small time steps at the beginning of the simulation. It furnishes outputs that focus only on the salient features of the response. The efficiency of the technique is demonstrated through the illustration of the results of three examples. Copyright © 2007 John Wiley & Sons, Ltd.

This paper described a technique for obtaining three-dimensional mine design information using a two-dimensional finite element program where the mining geometry consists of an extensive array of underground rooms and pillars. The technique is based upon a simple augmentation of forces in a two-dimensional analysis to produce the same average pillar stress that would occur in a full three dimensional analysis. Detailed comparisons between a three-dimensional analysis, a two-dimensional analysis (plane stress and plane strain) and an augmented two-dimensional analysis (also plane stress and plane strain) of stress about a typical coal mine pillar are presented. A local factor of safety is defined and then mapped over the pillar midplane, the immediate roof and immediate floor using the results from the full three-dimensional analysis. Comparisons of roof and pillar safety factor distributions obtained by the three-dimensional, two-dimensional and augmented two-dimensional analyses show that the minimum safety factors in the pillar (at the pillar sides) are predicted quite closely by the augmented two-dimensional techniqe (plane stress). The same is true of the immediate roof, although the three-dimensional safety factor tends to be higher in the roof (over the room) than that calculated by the augmented twodimensional technique. The augmented loading procedure appears to hold considerable promise as a very efficient and cost reducing techniqe for mine pillar design.

This work focuses on an analysis of dry joint retaining structures based on yield design theory: the stability of the masonry is assessed using rigid block and shear failure mechanisms in the wall and its backfill. An application of this simulation on 2D scale-down brick and wood models is then addressed, showing close agreement between theoretical predictions and experimental results. Further development on this work, including application of this theory on dry-stone retaining walls, is discussed as a conclusion. Copyright © 2009 John Wiley & Sons, Ltd.

This paper presents 2D and 3D upper bound solutions for the problem of tunnel excavation in soft ground. The solution invokes the use of incompressible flow fields derived from the theory of elasticity and the concept of sinks and sources. Comparison is made with previously published results. For some geometries the current calculation results in lower (better) upper bound values; however, the results were generally close to previously published values. Copyright (c) 2007 John Wiley & Sons, Ltd.

In this paper a coupled finite and boundary element formulation is developed for the analysis of excavation in jointed rock. The presence of joints in the rock mass has been included implicitly by treating it as an appropriate anisotropic elastic continuum. The boundary element formulation for an anisotropic medium is briefly discussed. Good agreement has been found between numerical and analytical solutions for several example problems, demonstrating the accuracy of the present formulation. Numerical solutions are also presented for the problems of a deep circular tunnel and a basement excavated in a variety of jointed rock masses.

A three-dimensional finite element simulation model for shield-driven tunnel excavation is presented. The model takes into account all relevant components of the construction process (the soil and the ground water, the tunnel boring machine with frictional contact to the soil, the hydraulic jacks, the tunnel lining and the tail void grouting). The paper gives a detailed description of the model components and the stepwise procedure to simulate the construction process. The soil and the grout material are modelled as saturated porous media using a two-field finite element formulation. This allows to take into account the groundwater, the grouting pressure and the fluid interaction between the soil and slurry at the cutting face and between the soil and grout around the tail void. A Cam-Clay plasticity model is used to describe the material behaviour of cohesive soils. The cementitious grouting material in the tail void is modelled as an ageing elastic material with time-dependent stiffness and permeability. To allow for an automated computation of arbitrarily long and also curvilinear driving paths with suitable finite element meshes, the simulation procedure has been fully automated. The simulation of a tunnel advance in soft cohesive soil below the ground water table is presented and the results are compared with measurements taken from the literature. Copyright © 2004 John Wiley & Sons, Ltd.

A systematic calibration procedure for the constitutive model gUTS using conventional triaxial data for all but one of the material constants is described. Typical ranges for the constants for clay and sand are specified together with default values. When all default values are adopted, just eight material constants need to be determined (gUTS-lite). A comprehensive series of 49 simulations on clays, silt and sands in loose and dense states under a wide range of monotonic and cyclic stress paths, initial states and drainage conditions provide very satisfactory agreement with experimental results.

A direct boundary element method for the iterative analysis of the lowered groundwater level and the steady-state airflow in porous soil for tunnels driven under compressed air is presented. The soil may be zoned and anisotropic. It is shown that disregard of the compressibility of the air leads to results for the excess air pressure and the flow of air through the surface of the soil, which are on the unsafe side. The lowered groundwater level is determined by means of an iterative procedure. During the iteration large changes of the shapes of boundary elements may occur. In order to reduce the resulting danger of divergence of the iteration, the boundary element mesh is adapted suitably in the course of the iteration process.

A modified three-dimensional discontinuous deformation analysis (3D-DDA) method is derived using four-noded tetrahedral elements to improve the accuracy of current 3D-DDA algorithm in practical applications. The analysis program for the modified 3D-DDA method is developed in a C++ environment and its accuracy is illustrated through comparisons with several analytical solutions that are available for selected problems. The predicted solutions for these problems using the modified 3D-DDA approach all show satisfactory agreement with the corresponding analytical results. Results presented in this paper demonstrate that the modified 3D-DDA method with discontinuous modeling capabilities offers a useful computational tool to determine stresses and deformations in practical problems involving fissured elastic media with reasonable accuracy. Copyright © 2008 John Wiley & Sons, Ltd.

This paper introduces a new generator algorithm and computer program for 3-D numerical simulation of packing configuration in a granular assemblies composed of ellipsoidal particles of different a/b aspect ratios. Each ellipsoidal particle is approximated by the revolution of an ellipse, formed by four connected arcs, about the major axis passing through its centroid. The centroid co-ordinates, major axis direction and lengths of the major and minor axes are the essential data for the packing generation and associated contact detection. The domain to be filled with particles can be a polyhedron of any shape. The packing program was coded based on a newly proposed scheme which obeys the no interpenetration kinematics of solid bodies.New contact detection algorithms for any two ellipsoids in the packing space were developed. Though simple, these algorithms effectively determine the contact condition and contact point without solving the simultaneous equations of the two ellipsoidal surfaces. Each particle's packing location, contact-point co-ordinates, and three-dimensional graphs can be created using the packing domain given boundaries, along with numbers, and geometrical information of particles to be generated.Simulation results show that this new algorithm provides an effective packing model as a required initial input for analysing the mechanics of granular material. This generation scheme potentially can explore the complex 3-D behaviours of material composed of discrete particles. Copyright © 1999 John Wiley & Sons, Ltd.

This paper presents a non-linear interface element to compute soil–structure interaction (SSI) based on the macro-element concept. The particularity of this approach lies in the fact that the foundation is supposed to be infinitely rigid and its movement is entirely described by a system of global variables (forces and displacements) defined in the foundation's centre. The non-linear behaviour of the soil is reproduced using the classical theory of plasticity. Failure is described by the interaction diagram of the ultimate bearing capacity of the foundation under combined loads. The macro-element is appropriate for modelling the cyclic or dynamic response of structures subjected to seismic action. More specifically, the element is able to simulate the behaviour of a circular rigid shallow foundation considering the plasticity of the soil under monotonic static or cyclic loading applied in three directions. It is implemented into FedeasLab, a finite element Matlab toolbox. Comparisons with experimental monotonic static and cyclic results show the good performance of the approach. Copyright

Rigid particle models taking directly into consideration the physical mechanisms and the influence of the material meso-structure have recently been developed for fracture studies of quasi-brittle material such as concrete. The formulation of a generalized contact model for rigid particle simulations is presented in which the contact discretization is a model parameter. The contact model performance for different discretizations is evaluated for uniaxial tensile tests, for uniaxial compression tests and for a notched beam in mode I. Copyright © 2005 John Wiley & Sons, Ltd.

This paper describes the development of a numerical model for compensation grouting which is a useful technique for the protection of surface structures from the potentially damaging movements arising from tunnel construction. Pipes are inserted into the ground between the tunnel and the overlaying structure from an access shaft. Buildings on the surface are instrumented and movements are carefully monitored. Once the deformations exceed a certain Trigger Level, grout is injected into the ground to prevent damage. In the finite element model described here, compensation grouting is modelled by applying an internal pressure to zero-thickness interface elements embedded in the mesh. An ‘observational algorithm’ is used, where the deformations of the surface are monitored and used to control the injection process. Example analyses of compensation grouting are given for three-dimensional tunnel construction underneath a greenfield site. Different strategies are used to control the injection process and their effectiveness in preventing surface movement is assessed. The numerical model is shown to replicate general behaviour expected in the field and is capable of modelling the control of ground surface movements at a greenfield site. Copyright © 2005 John Wiley & Sons, Ltd.

The solution of the poroelastic equations for predicting land subsidence above productive gas/oil fields may be addressed by the principle of virtual works using either the effective intergranular stress, with the pore pressure gradient regarded as a distributed body force, or the total stress incorporating the pore pressure. In the finite element (FE) method both approaches prove equivalent at the global assembled level. However, at the element level apparently the equivalence does not hold, and the strength source related to the pore pressure seems to generate different local forces on the element nodes. The two formulations are briefly reviewed and discussed for triangular and tetrahedral finite elements. They are shown to yield different results at the global level as well in a three-dimensional axisymmetric porous medium if the FE integration is performed using the average element-wise radius. A modification to both formulations is suggested which allows to correctly solve the problem of a finite reservoir with an infinite pressure gradient, i.e. with a pore pressure discontinuity on its boundary. Copyright © 2001 John Wiley & Sons, Ltd.

The stress variation induced by gas/oil production may activate pre-existing regional faults. This may enhance the expected land subsidence due to the generation of mechanically weak points close to the producing field. A class of elasto-plastic interface elements (IE), specifically designed to address the mechanical behaviour of faults over a regional scale, is integrated into a finite element (FE) geomechanical model and used to investigate the role exerted by active faults in anthropogenic land subsidence. The importance of regional faults depends on a variety of factors including depth of the depleted reservoir, fault number, orientation and size, geomechanical properties of porous medium, pore pressure drawdown induced by fluid production, etc. With the aid of some representative examples, a useful indication is provided as to where and how fault activation may influence both magnitude and extent of the land subsidence bowl above producing gas/oil reservoirs, pointing to a generally limited impact on the ground surface. The simulation of a real faulted gas reservoir in a complex 3-D setting shows that the proposed IE can be simply and efficiently incorporated into a FE geomechanical model, thus improving the quality of the stress and displacement prediction. Copyright © 2007 John Wiley & Sons, Ltd.

The surface subsidence above a compacting saturated oil reservoir is the main topic of this paper. From a literature review, it is obvious that extensive efforts have been conducted for investigating this phenomenon in various situations. Herein, a numerical model, based on the finite element method, was used for simulating three-dimensional three-phase fluid flow in a deforming saturated oil reservoir. The mathematical formulation describes a fully coupled governing equation system which consists of the equilibrium and continuity equations for three immiscible fluids flowing in a porous media. An elastoplastic soil model, based on a Mohr Coulomb yield surface, was utilized. The finite element method was applied to obtain simultaneous solutions to the governing equations where the displacements and the fluid pressures are the primary unknowns. The final discretized equations are solved by a direct solver using fully implicit procedures. A linear analysis was used to study the stability conditions of the present model. Finally, a series of simulations were conducted to indicate the validity and the utility of the developed model.

A linear boundary element (BE) model is proposed for the uncoupied simulation of land subsidence due to gas, oil and hot water production over three-dimensional (3-D) arbitrarily shaped reservoirs. The pore pressure decline is assumed to be specified in advance, e.g. via a numerical model of flow. Use is made of the fundamental solution derived in 1885 by Boussinesq for a vertical load acting upon the traction-free surface of a semi-infinite medium. A straightforward application of Betti's (1872) reciprocal theorem allows for the development of a boundary integral whose numerical execution yields directly the downward settlement over the point of interest. The new procedure is applied to assess land sinking caused by an uniform pore pressure decline occurring within fields of elliptical shape and to explore the influence of the assumption of small reservoir thickness which underlies the ‘tension center’ or ‘strain nucleus’ approach previously developed by Geertsma in 1966. The results emphasize the numerical efficiency of the solution and the promising features of the BE method for the evaluation of ground subsidence in 3-D problems. The present model is based on the theory of the linear poroelasticity and is implemented for a mechanically homogeneous and isotropic half-space. It allows for any arbitrary geometry of the reservoir and for a non-uniform distribution of the pore pressure decline. It may easily be extended to other physical settings for which a vertical surface point load solution is available.

In this paper, a model for the analysis of the behaviour of concrete at temperature largely exceeding critical point of water, is presented. In this temperature range liquid phase, i.e. capillary phase, and gas phase cannot be distinguished and only the latter exists. Consequently, capillary pressure has no more physical meaning above this point and liquid water is present only as physically adsorbed water. In this work, we give a different physical interpretation to the capillary pressure and use it still for the description of the hygrometric state of concrete in the zone, where temperature exceeds the critical point of water. Considerable thermal dilatation of the liquid water and the real behaviour of water vapour close to critical temperature are taken into account. Moreover, a special switching procedure in order to avoid the Stefan-like problem, which subsequently arises, is described and employed in the calculations. Finally, several numerical examples demonstrating the robustness of the adopted solution have been shown. Copyright © 2002 John Wiley & Sons, Ltd.

This paper assesses the performance of two commonly used absorbing boundaries in dynamic finite element analysis of geotechnical problems in conjunction with the domain reduction method (DRM). The DRM was originally developed by Bielak et al. (Bull. Seismol. Soc. Am. 2003; 93(2):817–824) to reduce the computational cost of seismological applications, whereas Yoshimura et al. (Bull. Seismol. Soc. Am. 2003; 93(2):825–840) showed that it can be effectively used as a boundary condition. In this study a practical methodology is proposed, which employs the cone boundary of Kellezi (Soil Dyn. Earthq. Eng. 2000; 19:533–547) on the outer boundary of the reduced (step II) model of the DRM. To verify the applicability of the proposed methodology, the results using both the cone boundary and the standard viscous boundary are compared with those using an extended mesh. Finally, results using the DRM as a boundary condition are compared with those using conventional boundary conditions. Some common pitfalls in the use of absorbing boundaries are highlighted and guidance for their correct use in engineering practice is given. Copyright © 2008 John Wiley & Sons, Ltd.

A non-linear seismic response analysis method for 2-D saturated soil–structure system with an absorbing boundary is presented. According to the 3-D strain space multimechanism model for the cyclic mobility of sandy soil, a constitutive expression for the plane strain condition is first given. Next, based on Biot's two-phase mixture theory, the finite element equations of motion for a saturated soil–structure system with an absorbing boundary during earthquake loadings are derived. A simulation of the shaking table test is performed by applying the proposed constitutive model. The effectiveness of the absorbing boundary is examined for the 2-D non-linear finite element models subjected to random inputs. Finally, a numerical seismic response analysis for a typical saturated soil–structure system is performed as an application of the proposed method.

Water-absorbing rocks are formed from minerals that can hold water in their crystal structure or between grain boundaries. Such water absorption is often accompanied by a change in the crystal dimension that manifests itself as a swelling of the rock. Swelling is particularly pronounced in rocks containing phyllosilicates because of the ease with which these minerals hydrate; it is thus of geological and geotechnical relevance in shales, clay-rich soils and zeolitized tuffs. The model of hydration swelling that we present here is based on extended versions of the equations of poroelasticity and Darcy's transport law, which we derive using a non-equilibrium thermodynamics approach. Our equations account for the hydration reaction under the assumption that the reaction rate is fast in comparison with the rate at which hydraulic state changes are communicated through the rock, i.e. that local physico-chemical equilibrium persists. Using a finite-element scheme for solving numerically the governing equations of our model, we simulate the creep of shales during a routine swelling test and calculate the stress and strain distributions around wellbores drilled in shale formations that undergo swelling. We show that swelling effects promote tensile failure of the wellbore wall.

In this paper, time-domain dynamic analysis of dam–reservoir–foundation interaction is presented by coupling the dual reciprocity boundary element method (DRBEM) in the infinite reservoir and foundation domain and the finite element method in the finite dam domain. An efficient coupling procedure is formulated by using the sub-structuring method. The effects of the reservoir bottom absorption are included in the formulations. Sharan's boundary condition for the far-end of the infinite fluid domain is implemented. To verify the proposed scheme, numerical examples are carried out by comparing with the available exact solutions and finite–finite element coupling results for the dam–reservoir interaction. A complete dam–reservoir–foundation interaction is also studied by including the bottom absorption effects. Copyright © 2004 John Wiley & Sons, Ltd.

The parameter identification procedure proposed in this paper is based on the solution of an inverse problem, which relies on the minimization of an error function of least-squares type. The solution of the ensuing optimization problem, which is a constrained one owing to the presence of physical links between the optimization parameters, is performed by means of a particular technique of the feasible direction type, which is modified and improved when the problem turns to an unconstrained one. The algorithm is particularly efficient in the presence of hierarchical material models. The numerical properties of the proposed procedure are discussed and its behaviour is compared with usual optimization methods when applied to constrained and unconstrained problems. Copyright © 2001 John Wiley & Sons, Ltd.

The most flexible and generally applicable methods for elasto-plastic analysis are those based on an incremental-iterative form of the initial stress approach, but such methods often exhibit slow convergence. The acceleration procedure known as the alpha-constant stiffness method is reconsidered and some modifications are proposed. The principal difference in the present approach lies in the use of a single acceleration parameter, rather than a diagonal matrix of acceleration coefficients. The new scheme shows a significant improvement in numerical stability and converges three times faster than the standard initial stress method. Some practical aspects associated with the method are discussed and a number of applications are presented.

Most reinforced concrete structures are damaged due to corrosion of reinforcements in concrete. In normal conditions the pH near the reinforcements is around 12–13 which means that steel is in a passive state. But aggressive species, such as chloride ions or carbon dioxide, may penetrate into concrete and promote active corrosion. As a consequence (hydro)oxides are produced leading to degradation of concrete structures. For instance cracking of the concrete is generated due to the pressure induced by rust. In this paper, we study the inception and the propagation of cracking on reinforced mortar plates with rebars located either in the middle or at the corner. Additional experiments have been performed on cylindrical specimens to determine the local effect of rust pressure at the interface rust/mortar. The specimens have been subjected to imposed current density in order to enhance the corrosion and digital image intercorrelation has been used to determine displacement fields. The experiments have been compared to numerical modelling. Copyright © 2010 John Wiley & Sons, Ltd.

In this paper, a novel combination of well-established numerical procedures is explored in order to accelerate the simulation of sequential excavation. Usually, large-scale models are used to represent these problems. Due to the high number of equations involved, the solver algorithm represents the critical aspect which makes the simulation very time consuming. The mutable nature of the excavation models makes this problem even more pronounced. To accomplish the representation of geometrical and mechanical aspects in an efficient and simple manner, the proposed solution employs the boundary element method with a multiple-region strategy. Together with this representational system, a segmented storage scheme and a time-ordered tracking of the changes form an adequate basis for the usage of fast updating methods instead of frontal solvers. The present development employs the Sherman–Morrison–Woodbury method to speed up the calculation due to sequential changes. The efficiency of the proposed framework is illustrated through the simulation of test examples of 2D and 3D models. Copyright © 2006 John Wiley & Sons, Ltd.

The pressure variations during the production of petroleum reservoir induce stress changes in and around the reservoir. Such changes of the stress state can induce marked deformation of geological structures for stress sensitive reservoirs as chalk or unconsolidated sand reservoirs. The compaction of those reservoirs during depletion affects the pressure field and so the reservoir productivity. Therefore, the evaluation of the geomechanical effects requires to solve in a coupling way the geomechanical problem and the reservoir multiphase fluid flow problem. In this paper, we formulate the coupled geomechanical-reservoir problem as a non-linear fixed point problem and improve the resolution of the coupling problem by comparing in terms of robustness and convergence different algorithms. We study two accelerated algorithms which are much more robust and faster than the conventional staggered algorithm and we conclude that they should be used for the iterative resolution of coupled reservoir-geomechanical problem. Copyright © 2006 John Wiley & Sons, Ltd.

The purpose of this paper is to examine the importance of different possible simplifying approximations when performing numerical simulations of fluid-filled porous media subjected to dynamic loading. In particular, the relative importance of the various acceleration terms for both the solid and the fluid, especially the convective contribution, is assessed. The porous medium is modelled as a binary mixture of a solid phase, in the sense of a porous skeleton, and a fluid phase that represents both liquid and air in the pores. The solid particles are assumed to be intrinsically incompressible, whereas the fluid is assigned a finite intrinsic compressibility. Finite element (FE) simulations are carried out while assuming material properties and loading conditions representative for a road structure. The results show that, for the range of the material data used in the simulations, omitting the relative acceleration gives differences in the solution of the seepage velocity field, whereas omitting only the convective term does not lead to significant differences. Copyright © 2007 John Wiley & Sons, Ltd.

A new and simple concept based on the idea of correcting for non-associative characteristics of (geologic) media is presented. A special form of the concept is adopted and introduced in a critical state plasticity model. An example problem of behaviour of a soil tested under triaxial conditions is included. The concept can permit a simplified treatment of non-associativeness and under certain assumptions can allow use of existing formulations of plasticity by maintaining symmetry of the associated matrices.

In this work, we present a numerical procedure for determining the nature stress state in the rock mass around a tunnel. A finite element method is applied for analyzing the direct problems of tunneling during the back analysis of parameter estimation, in which a no-tension elastic–plastic model is used to simulate the elastic–tensile and elastic–plastic-tensile failure states which often occur in the cases of underground excavation in heavily jointed rock masses. By considering the natural stress state as random parameters of the tunneling system, the Kalman filter method is employed for feedback analysis to modify the parameter values in a statistical context, which uses the prior information in the process of estimation and employs a set of displacements obtained from field measurements. To verify the effectiveness of the proposed method of inverse analysis, the developed numerical procedure is applied to a synthetic example of deep tunnels in yielding rock masses. The relative importance of the a priori and updating information is investigated, as is the importance of their uncertainty. The results show great potential of the proposed approach. Copyright © 2010 John Wiley & Sons, Ltd.

This paper prepares the ground for the continuum analysis of shear band evolution using a Cosserat/micropolar constitutive equation derived from micromechanical considerations. The nature of the constitutive response offers two key advantages over other existing models. Firstly, its non-local character obviates the mathematical difficulties of traditional analyses, and facilitates an investigation of the shear band evolution (i.e. the regime beyond the onset of localization). Secondly, the constitutive model parameters are physical properties of particles and their interactions (e.g. particle stiffness coefficients, coefficients of inter-particle rolling friction and sliding friction), as opposed to poorly understood fitting parameters. In this regard, the model is based on the same material properties used as model inputs to a discrete element (DEM) analysis, therefore, the micromechanics approach provides the vehicle for incorporating results not only from physical experiments but also from DEM simulations. Although the capabilities of such constitutive models are still limited, much can be discerned from their general rate form. In this paper, an attempt is made to distinguish between those aspects of the continuum theory of localization that are independent of the constitutive model, and those that require significant advances in the understanding of micromechanics. Copyright © 2004 John Wiley & Sons, Ltd.

Repeated small amplitude dynamic loading of the soil in the vicinity of buildings, as arising from traffic or construction activities, may cause differential foundation settlements and structural damage. In this paper, a numerical model for soils under repeated dynamic loading is formulated. It is assumed that the dynamic part of the loading is small with respect to the static part, reflecting the stress conditions in the soil underneath buildings. As the plastic deformation in the soil is only observed after a considerable amount of dynamic loading cycles, only the accumulation of the average plastic deformation is considered. The model accounts for the dependency of the deformation on the stress conditions and the dynamic loading amplitude. The accumulation model is implemented in a finite element framework, using a consistent tangent approach in combination with a backward Euler integration scheme. A triaxial test is considered in a first numerical example. The available analytical solution for this problem allows to validate the numerical implementation. Second, the differential settlement of a two-storey building founded on loose sandy soil under repeated vehicle passages is considered. The differential foundation settlement causes the stresses to increase at the bottom of the wall, which may result in damage. Copyright © 2009 John Wiley & Sons, Ltd.