Dynamic impedances of pile groups with embedded caps in homogeneous elastic soils using CIFECM

Significant research has been reported on the dynamic analysis of pile groups. However, in most of the cases, the effect of pile cap is neglected despite the fact that there may be additional interactions due to the presence of the cap. This paper presents the dynamic impedances for the pile groups with caps embedded in isotropic homogeneous elastic soils. A general three-dimensional finite element procedure is developed. The system is sub-structured into bounded near-field and an unbounded far-field. The pile-soil system of the near-field is modeled using solid finite elements, and the unbounded elastic soil system of the far-field is modeled using the consistent infinitesimal finite element cell method (CIFECM) in the frequency domain.     

Two-dimensional development of the dynamic coupled consolidation scaled boundary finite-element method for fully saturated soils

The scaled boundary finite-element method is a powerful tool used to analyse far-field boundary soil–structure interaction problems. In this paper, the method is extended to include Biot's coupled consolidation in order to deal with fully saturated soil as a two-phase medium. The advantages of this method are explained in this paper. The detailed formulation considers the general two-dimensional (2D) analysis case, accounting for body forces and surface tractions in both bounded and unbounded media.     

One-dimensional modelling of the non-linear far field in soil–structure-interaction analysis

In a bounded domain elasto-plastic wave propagation can be modelled accurately using the finite-element method. As is even the case for an elastic analysis, an unbounded domain, e.g. a semi-infinite soil or fluid, can, however, not be represented in this manner, as any spatial discretization has to be avoided. For one-dimensional wave propagation with a bi-linear elasto-plastic material law involving one stress component an analytical solution exists. The latter is used in modelling the non-linear far field of an unbounded medium using a rigorous bookkeeping procedure of the generated elastic and plastic waves propagating in both directions. The need for a non-linear model of the far field arises, as in a two-dimensional representation of soil-structure interaction the surface waves do not decay.     

Dynamic stiffness of unbounded medium based on damping-solvent extraction

In the damping-solvent extraction method, to calculate the dynamic-stiffness matrix of an unbounded medium, a finite region of the medium, adjacent to the structure is analysed in the first step, whereby hysteretic material damping is introduced artificially as a solvent. This leads to the dynamic-stiffness matrix of the damped bounded medium, which is assumed in the second step to be equal to that of the damped unbounded medium. In the third step, the effect of the material damping on the dynamic-stiffness matrix is eliminated, i.e. the damping solvent is extracted, resulting in the dynamic-stiffness matrix of the unbounded medium. The damping-solvent extraction method permits an efficient calculation of the dynamic-stiffness matrix of an unbounded medium by analysing the adjacent bounded medium only, which exhibits the same dynamic characteristics as the (bounded) structure. The familiar standard finite-element method is sufficient for the analysis and the hysteretic damping is introduced by multiplying the elastic moduli by 1 + 2i£. The introduced hysteretic material damping, the solvent, is extracted at the end of the analysis for each coefficient of the dynamic-stiffness matrix and for each frequency independently of the others by a very concise equation based on a Taylor expansion. The method is evaluated thoroughly for dynamic soil-structure interaction and for seismic reservoir-dam interaction using stringent simple cases with analytical solutions available and is also applied to practical examples, by calculating the dynamic-stiffness matrix of a semi-infinite wedge and an embedded foundation.     

The near-field method: a modified equivalent linear method for dynamic soil–structure interaction analysis. Part I: Theory and methodology

This paper presents a new direct modeling approach to analyze 3D dynamic SSI systems including building structures resting on shallow spread foundations. The direct method consists of modeling the superstructure and the underlying soil domain. Using a reduced shear modulus and an increased damping ratio resulted from an equivalent linear free-field analysis is a traditional approach for simulating behavior of the soil medium. However, this method is not accurate enough in the vicinity of foundation, or the near-field domain, where the soil experiences large strains and the behavior is highly nonlinear. This research proposes new modulus degradation and damping augmentation curves for using in the near-field zone in order to obtain more accurate results with the equivalent linear method. The mentioned values are presented as functions of dimensionless parameters controlling nonlinear behavior in the near-field zone. This paper summarizes the semi-analytical methodology of the proposed modified equivalent linear procedure. The numerical implementation and examples are given in a companion paper.     

基础动力刚度的精确数值解及集中参数模型

土－结构相互作用分析的关键是建立以土－结构界面定义的无限半空间的动力刚度矩阵。本文介绍了一种求解半无限地基动力刚度的新方法，通过两个算例验证了该方法的精度，并给出了一种利用频域刚性基础动力刚度计算基础时域荷载响应的实用方法，该研究为刚性基础设计提供了一种新的，可靠的理论方法。     

Dynamic responses of structure–soil–structure systems with an extension of the equivalent linear soil modeling

A three-dimensional problem of cross interaction of adjacent structures through the underlying soil under seismic ground motion is investigated. The story shears and lateral relative displacements (drifts) are the targets of the computations. These are calculated using a detailed modeling of soil, the foundations and the two adjacent structures. An equivalent linear behavior is assumed for the soil by introducing reduced mechanical properties consistent with the level of ground shaking for the free-field soil. Then a distinctive soil zone (the near-field soil) is recognized in the vicinity of the foundations where the peak shear strain under the combined effect of a severe earthquake and the presence of structures is much larger than the strain threshold up to which the soil can be modeled as an equivalent linear medium. It is shown that it is still possible to use an equivalent linear behavior for the near-field soil if its shear modulus is further reduced with a factor depending on the dynamic properties of the adjacent structures, the near-field soil, and the design earthquake. Variations of the dynamic responses of different adjacent structures with their clear distances are also discussed.     

CONSISTENT INFINITESIMAL FINITE-ELEMENT CELL METHOD IN FREQUENCY DOMAIN

To calculate the dynamic-stiffness matrix at the structure–medium interface of an unbounded medium for the range of frequencies of interest, the consistent infinitesimal finite-element cell method based on finite elements is developed. The derivation makes use of similarity and finite-element assemblage, yielding a non-linear first-order ordinary differential equation in frequency. The asymptotic expansion for high frequency yields the boundary condition satisfying the radiation condition. In an application only the structure–medium interface is discretized resulting in a reduction of the spatial dimension by one. The boundary condition on the free surface is satisfied automatically. The consistent infinitesimal finite-element cell method is exact in the radial direction and converges to the exact solution in the finite-element sense in the circumferential directions. Excellent accuracy results.     

Consistent formulation of direct and substructure methods in nonlinear soil-structure interaction

Both direct and substructure methods of dynamic soil-structure interaction analysis can be treated using a common analytical model with difference being restricted only to the definition of boundary conditions of the bounded soil zone. It is shown that a consistent formulation of the problem equally applicable to both methods can be achieved in which true nonlinear behaviour of the bounded soil zone (near-field) can be taken into account in the time domain through properly defined constitutive models. However, for the linear boundary conditions to be imposed on the near-field inevitably involves the application of the principle of superposition resulting in a linear far-field approximation. Therefore, the bounded soil zone taken should be large enough in both methods to reduce the adverse effects of the far-field linearization.     

Dynamic-stiffness matrix in time domain of unbounded medium by infinitesimal finite element cell method

To calculate the dynamic-stiffness matrix in the time domain (unit-impulse response functions) of the unbounded medium, the infinitesimal finite element cell method based solely on the finite element formulation and working exclusively in the time domain is developed. As in the cloning algorithm, the approach is based on similarity of the unbounded media corresponding to the interior and exterior boundaries of the infinitesimal finite element cell. The derivation can be performed exclusively in the time domain, or alternatively in the frequency domain. At each time station a linear system of equations is solved. The consistent-boundary method to analyse a layered medium in the frequency domain and the viscous-dashpot boundary method are special cases of the infinitesimal finite element cell method. The error is governed by the finite element discretization in the circumferential direction, as the width of the finite-element cell in the radial direction is infinitesimal. The infinitesimal finite element cell method is thus ‘exact in the finite-element sense’. This method leads to highly accurate results for a vast class of problems, ranging from a one-dimensional spherical cavity to a rectangular foundation embedded in a half-plane.     

Response of unbounded soil in scaled boundary finite‐element method

The scaled boundary finite‐element method is a powerful semi‐analytical computational procedure to calculate the dynamic stiffness of the unbounded soil at the structure–soil interface. This permits the analysis of dynamic soil–structure interaction using the substructure method. The response in the neighbouring soil can also be determined analytically. The method is extended to calculate numerically the response throughout the unbounded soil including the far field. The three‐dimensional vector‐wave equation of elasto‐dynamics is addressed. The radiation condition at infinity is satisfied exactly. By solving an eigenvalue problem, the high‐frequency limit of the dynamic stiffness is constructed to be positive definite. However, a direct determination using impedances is also possible. Solving two first‐order ordinary differential equations numerically permits the radiation condition and the boundary condition of the structure–soil interface to be satisfied sequentially, leading to the displacements in the unbounded soil. A generalization to viscoelastic material using the correspondence principle is straightforward. Alternatively, the displacements can also be calculated analytically in the far field. Good agreement of displacements along the free surface and below a prism foundation embedded in a half‐space with the results of the boundary‐element method is observed. Copyright © 2001 John Wiley & Sons, Ltd.     

Non-linear seismic response analysis of soil-structure interaction systems

This paper presents an effective analysis procedure for the dynamic soil-structure interaction problem considering not only the sliding and separation phenomena but also the non-linear behaviour of soil by the finite element method. Soil is assumed to be an elasto-plastic material and the contact surface between the soil and structure is modelled by the joint element. The load transfer method is adopted to carry out dynamic non-linear response analysis. The method is applied to the response analysis of a nuclear reactor building resting on the ground surface. The effects of non-linear behaviour of soil on the safety against sliding of the structure are examined. The numerical computations reveal the following results: that the non-linear behaviour of soil reduces the response of the system and the magnitude of sliding of the structure, and that the safety against sliding obtained by the proposed method is higher than the safety obtained by classical methods. This implies the possibility of a more rational and economical design of large structures; it can be said that the proposed method provides useful information for the stability analysis of important and large structures.     

Soil-structure interaction using BEM–FEM coupling through ANSYS software package

The main aim of this work is to develop, verify and apply in simulation study an efficient hybrid approach to study seismic response of a soil-structure system taking into account all the important components as: (1) the line time-harmonic source with its specific geophysical properties; (2) the inhomogeneity and heterogeneity of the wave path from the source to the local geological region; (3) the geotechnical properties of the near-field local geological profile and finally (4) the properties of the engineering structure itself. Plane strain state is considered. The hybrid computational tool is based on the boundary element method (BEM¹) for modeling the infinite far-field geological media and finite element method (FEM²) for treating the dynamic behavior of the structure and the near-field finite soil geological region. Each of the two techniques is applied in that part of the whole model where it works more efficiently. The hybrid numerical scheme is realized via the sub-structure approach, direct BEM¹, conventional FEM² and insertion of the BEM¹ model of the seismically active far-field geological media as a macro-finite element (MFE³) in the FEM² commercial program ANSYS. The accuracy and verification study of the proposed method is presented by solution of numerical test examples simulating different seismic scenarios. The obtained results show clearly that the hybrid model is able to demonstrate the sensitivity of the synthetic signals to the source properties, to the heterogeneous character of the wave path, to the relief peculiarities of the local layered geological deposit and to the specific properties of the engineering structure.     

饱和土体瞬态响应有限元分析

给出基于Biot多孔介质理论分析饱和土体在动载荷作用下瞬态响应的有限元公式,数值计算部分采用本文有限元法分别计算一维饱和土柱在两种不同类型动载荷作用下的瞬态响应,并将数值计算结果与文献中的解析解进行比较,二者结果十分吻合,从而验证本文方法的可行性。     

The near-field method: a modified equivalent linear method for dynamic soil–structure interaction analysis. Part II: verification and example application

Usually for modeling of soil in a direct soil–structure interaction (SSI) problem, the equivalent linear soil properties are used. However, this approach is not valid in the vicinity of a foundation, where the soil experiences large strains and a high level of nonlinearity because of structural vibrations. The near-field method was developed and described in a companion paper to overcome this limitation. This method considers the effects of large strains and suggests a shear modulus and a damping ratio further modified in the near-field of a foundation. Validity and performance of this approach are evaluated, application examples are explained and the results of a parametric study about the role of soil and structure parameters in the extent of SSI effects on the nonlinear seismic response of structures are presented in this paper. One real existing and five, ten, fifteen and twenty story moment-resisting frame steel buildings with two different site conditions corresponding to firm and soft soils are considered and the responses obtained from the near-field method are compared with the recorded and rigorous responses. Moreover, various SSI modeling techniques are employed to investigate the accuracy and performance of each approach. The results show that the near-field method is a simple yet accurate enough approach for analysis of direct SSI problems.     

Three-dimensional dynamic soil–structure interaction analysis in the time domain

A new numerical procedure is proposed for the analysis of three-dimensional dynamic soil–structure interaction in the time domain. In this study, the soil is modelled as a linear elastic solid, however, the methods developed can be adapted to include the effects of soil non-linearities and hysteretic damping in the soil. A substructure method, in which the unbounded soil is modelled by the scaled boundary finite-element method, is used and the structure is modelled by 8–21 variable-number-node three-dimensional isoparametric or subparametric hexahedral curvilinear elements. Approximations in both time and space, which lead to efficient schemes for calculation of the acceleration unit-impulse response matrix, are proposed for the scaled boundary finite-element method resulting in significant reduction in computational effort with little loss of accuracy. The approximations also lead to a very efficient scheme for evaluation of convolution integrals in the calculation of soil–structure interaction forces. The approximations proposed in this paper are also applicable to the boundary element method. These approximations result in an improvement over current methods. A three-dimensional Dynamic Soil–Structure Interaction Analysis program (DSSIA-3D) is developed, and seismic excitations (S-waves, P-waves, and surface waves) and externally applied transient loadings can be considered in analysis. The computer program developed can be used in the analysis of three-dimensional dynamic soil–structure interaction as well as in the analysis of wave scattering and diffraction by three-dimensional surface irregularities. The scattering and diffraction of seismic waves (P-, S-, and Rayleigh waves) by various three-dimensional surface irregularities are studied in detail, and the numerical results obtained are in good agreement with those given by other authors. Numerical studies show that the new procedure is suitable and very efficient for problems which involve low frequencies of interest for earthquake engineering. Copyright © 1999 John Wiley & Sons Ltd     

A comparison of time-domain transmitting boundaries

A non-linear interaction analysis with a (generalized) non-linear structure and a linear unbounded soil is analysed in the time domain, based either on the sub-structure method, which involves global convolution integrals, or on the direct method with local boundary conditions. Alternatively, the hybrid frequency–time-domain method of analysis, which is an iterative scheme, could be used. Approximate local boundary conditions to model the wave propagation towards infinity on the artificial boundary used in the direct method of non-linear soil–structure-interaction analysis to be performed in the time domain are examined. A semi-infinite rod supported elastically, which exhibits the same properties as certain unbounded soils such as dispersion and a cut-off frequency, is used for the investigation. For a transient excitation, the superposition boundary with frequent averaging, the well-known viscous damper and the extrapolation algorithm lead to good accuracy. Moving the artificial boundary further away from the structure (or more precisely, increasing the ratio of the distance of the artificial boundary to the wave length) improves the accuracy.     

Identification of the Hualien soil-structure interaction system

This article demonstrates how system identification techniques can be successfully applied to a soil-structure interaction system in conjunction with the results of the forced vibration tests on the Hualien large-scale seismic test structure which was recently built in Taiwan for an international joint research. The parameters identified are the shear moduli of several near-field soil regions as well as Young's moduli of the shell sections of the structure. The soil-structure interaction system is represented by the finite element method combined with infinite element formulation for the unbounded layered soil medium. Preliminary investigations are carried out on the results of the static stress analysis for the soil medium and the results of the in-situ tests to divide the soil-structure system into several regions with homogeneous properties and to determine the lower and upper bounds of the parameters for the purpose of identification. Then two sets of parameters are identified for two principal directions based on the forced vibration test data by minimizing the estimation error using the constrained steepest descent method. The simulated responses for the forced vibration tests using the identified parameters show excellent agreement with the test data. The present estimated parameters are also found to be well compared with the average value of those by other researchers in the joint project.