A meshless local boundary integral equation method for solving transient elastodynamic problems

A new local boundary integral equation (LBIE) method for solving two dimensional transient elastodynamic problems is proposed. The method utilizes, for its meshless implementation, nodal points spread over the analyzed domain and employs the moving least squares (MLS) approximation for the interpolation of the interior and boundary variables. On the global boundary, displacements and tractions are treated as independent variables. The local integral representation of displacements at each nodal point contains both surface and volume integrals, since it employs the simple elastostatic fundamental solution and considers the acceleration term as a body force. On the local boundaries, tractions are avoided with the aid of the elastostatic companion solution. The collocation of the local boundary/volume integral equations at all the interior and boundary nodes leads to a final system of ordinary differential equations, which is solved stepwise by the -Wilson finite difference scheme. Direct numerical techniques for the accurate evaluation of both surface and volume integrals are employed and presented in detail. All the strongly singular integrals are computed directly through highly accurate integration techniques. Three representative numerical examples that demonstrate the accuracy of the proposed methodology are provided.     

基于二项级数近似的结构概率分析

结构稳健优化设计中,一个关键的环节是分析结构响应量的概率特性,即计算响应的均值和方差。常用的方法主要有泰勒级数法、蒙特卡洛法以及数值积分法等。其中泰勒级数法精度较差,不适用于参数方差较大的随机结构,而蒙特卡洛法和高斯积分法计算量又过大。为了提高结构稳健性分析的计算效率,将结构位移的二项级数近似技术引入到高斯积分方法之中,提出一种结构位移均值及方差的计算方法。同时,用伴随向量法推导了相关的灵敏度计算公式。通过一个算例与已有的方法进行了比较,表明该方法较大程度上减少了高斯积分法的计算量,而与泰勒级数法相比,该方法又具有较高的计算精度,并且其灵敏度计算不再需要重分析,计算量较少。     

Enrichment of the method of finite spheres using geometry‐independent localized scalable bubbles

In this paper, we report the development of two new enrichment techniques for the method of finite spheres, a truly meshfree method developed for the solution of boundary value problems on geometrically complex domains. In the first method, the enrichment functions are multiplied by a weight function with compact support, while in the second one a floating ‘enrichment node’ is introduced. The scalability of the enrichment bubbles offers flexibility in localizing the spatial extent to which the enrichment field is applied. The bubbles are independent of the underlying geometric discretization and therefore provide a means of achieving convergence without excessive refinement. Several numerical examples involving problems with singular stress fields are provided demonstrating the effectiveness of the enrichment schemes and contrasting them to traditional ‘geometry‐dependent’ enrichment strategies in which one or more nodes associated with the geometric discretization of the domain are enriched. An additional contribution of this paper is the use of a meshfree numerical integration technique for computing the J‐integral using the domain integral method. Copyright © 2006 John Wiley & Sons, Ltd.     

求矩阵符号函数的割线法及其收敛性

符号函数是求解来自控制论中相关的Lyapunov方程和Riccati方程的有力工具,它也用来解某些特征值问题和计算不变子空间.本文给出了求矩阵符号函数的割线法,证明了该方法对于特殊的初始矩阵是全局超线性收敛的,并给出了数值试验,并将割线法与Newton法进行了比较,理论上和数值上均验证了割线法是求矩阵符号函数的有效数值方法.     

Axisymmetric nonlinear analysis of shallow spherical shells by a combined boundary element and finite difference method

This paper is concerned with the development of the mixed boundary element method and finite element method for the analysis of spherical annular shells under axisymmetric loads. The boundary element techniques are used to solve the equilibrium equation of shells and the central difference operator is adopted to deal with the compatibility equations. Iterative techniques are used throughout the analysis procedure. A number of numerical examples are given in the paper to illustrate the validity of the present approach.     

CQM-based BEM formulation for uncoupled transient quasistatic thermoelasticity analysis

A boundary element method based on the convolution quadrature method for the numerical solution of uncoupled transient thermoelasticity problems is presented. In the proposed formulation, the time-domain integral equation is numerically approximated by a quadrature formula whose weight factors are computed by means of integral expressions involving the Laplace transform of the fundamental solution. Compared with other numerical methods that operate directly in the Laplace transformed domain, the proposed formulation requires only the definition of the time-step used by the procedure of integration, and does not need special techniques of inversion from the Laplace-domain to the time-domain. Numerical examples of transient thermoelasticity problems are presented to show the versatility and accuracy of the method.     

Generalized finite difference method for solving two-dimensional non-linear obstacle problems

In this study, the obstacle problems, also known as the non-linear free boundary problems, are analyzed by the generalized finite difference method (GFDM) and the fictitious time integration method (FTIM). The GFDM, one of the newly-developed domain-type meshless methods, is adopted in this study for spatial discretization. Using GFDM can avoid the tasks of mesh generation and numerical integration and also retain the high accuracy of numerical results. The obstacle problem is extremely difficult to be solved by any numerical scheme, since two different types of governing equations are imposed on the computational domain and the interfaces between these two regions are unknown. The obstacle problem will be mathematically formulated as the non-linear complementarity problems (NCPs) and then a system of non-linear algebraic equations (NAEs) will be formed by using the GFDM and the Fischer–Burmeister NCP-function. Then, the FTIM, a simple and powerful solver for NAEs, is used solve the system of NAEs. The FTIM is free from calculating the inverse of Jacobian matrix. Three numerical examples are provided to validate the simplicity and accuracy of the proposed meshless numerical scheme for dealing with two-dimensional obstacle problems.     

Numerical solution of two backward parabolic problems using method of fundamental solutions

This work is devoted to a numerical algorithm based on the method of fundamental solutions (MFS) for solving two backward parabolic problems with different boundary conditions, one with nonlocal Dirichlet boundary conditions, and second one with Robin type boundary conditions. The initial temperature distribution will be identified from the final temperature distribution, which appear in some applied subjects. The Tikhonov regularization method with the L-curve criterion for choosing the regularization parameter is adopted for solving the resulting matrix equation which is highly ill-conditioned. Two numerical examples are provided to show the high efficiency of the suggested method.     

Hybridizing evolutionary strategies with continuation methods for solving multi-objective problems

Two techniques for the numerical treatment of multi-objective optimization problems—a continuation method and a particle swarm optimizer—are combined in order to unite their particular advantages. Continuation methods can be applied very efficiently to perform the search along the Pareto set, even for high-dimensional models, but are of local nature. In contrast, many multi-objective particle swarm optimizers tend to have slow convergence, but instead accomplish the ‘global task’ well. An algorithm which combines these two techniques is proposed, some convergence results for continuous models are provided, possible realizations are discussed, and finally some numerical results are presented indicating the strength of this novel approach.     

A discrete splitting finite element method for numerical simulations of incompressible Navier–Stokes flows

The presence of the pressure and the convection terms in incompressible Navier–Stokes equations makes their numerical simulation a challenging task. The indefinite system as a consequence of the absence of the pressure in continuity equation is ill‐conditioned. This difficulty has been overcome by various splitting techniques, but these techniques incur the ambiguity of numerical boundary conditions for the pressure as well as for the intermediate velocity (whenever introduced). We present a new and straightforward discrete splitting technique which never resorts to numerical boundary conditions. The non‐linear convection term can be treated by four different approaches, and here we present a new linear implicit time scheme. These two new techniques are implemented with a finite element method and numerical verifications are made. Copyright © 2005 John Wiley & Sons, Ltd.     

Generalized hybrid-combined methods for the singularity problems of homogeneous elliptic equations

Combinations of the Ritz–Galerkin and finite element methods are applied to solving singularity problems of homogeneous, elliptic equations. The Ritz-Galerkin method is used in the subdomains where there exist singular points; and the finite element method is still used in the rest of the solution domains. More general coupling techniques than those of Reference 6 along the common boundary of subdomains are discussed. Numerical experiments using these kinds of coupling techniques are provided for the first time. It is interesting that the calculated results of Motz's problem have shown the simplified hybrid strategy in Reference 6 to be optimal for both error bounds and stability of numerical solutions, among all general coupling techniques.     

On the numerical solution of non-linear reservoir flow models with gravity

Based on a two-phase fluid model for immiscible displacement in a porous medium, we develop and analyse numerical solution techniques for certain non-linear phenomena. Two different solution strategies for the treatment of gravity effects, which are non-trivial to model by existing solution techniques and may be of great influence in many practical flow situations, are presented. The solution procedures are based on an operator-splitting technique, combining the modified method of characteristics with finite element techniques and adaptive grid refinement.     

A 3D mortar method for solid mechanics

A version of the mortar method is developed for tying arbitrary dissimilar 3D meshes with a focus on issues related to large deformation solid mechanics. Issues regarding momentum conservation, large deformations, computational efficiency and bending are considered. In particular, a mortar method formulation that is invariant to rigid body rotations is introduced. A scheme is presented for the numerical integration of the mortar surface projection integrals applicable to arbitrary 3D curved dissimilar interfaces. Here, integration need only be performed at problem initialization such that coefficients can be stored and used throughout a quasi‐static time stepping process even for large deformation problems. A degree of freedom reduction scheme exploiting the dual space interpolation method such that direct linear solution techniques can be applied without Lagrange multipliers is proposed. This provided a significant reduction in factorization times. Example problems which touch on the aforementioned solid mechanics related issues are presented. Published in 2003 by John Wiley & Sons, Ltd.     

On the accurate assessment of crack opening and closing stresses in plasticity-induced fatigue crack closure problems

The accurate calculation of the opening and closing stresses is an important issue in fatigue crack closure problems, since the effective driving force for crack growth is dependent on accurate calculation of the opening stresses. Often numerical methods such as finite element analysis are used to model plasticity-induced fatigue crack closure problems. There are many difficulties associated with this modelling work, since the results may depend on a wide range of parameters such as mesh refinement, node release scheme and modelling of the contact between the crack faces etc. Even after a great deal of modelling work some arbitrariness is evident in the technique used for assessing the opening and closing stresses. A number of techniques have been proposed in the literature and the current work will assess and compare these approaches. The node displacement method, the change in stresses at the crack tip, and the weight function technique will each be applied to a finite element model of a plane stress crack for a range of stress levels. In addition, an analytical model for plasticity-induced crack closure under plane stress conditions will be used to discuss the accuracy of these techniques. The investigation shows that all these techniques are equivalent provided that the displacement and stress at the crack tip are assessed accurately. However, it will be shown that use of the tensile tip stress method, proposed by some authors for assessing the closing stress, is erroneous.     

Timoshenko梁能量守恒逐步积分算法

该文提出了Timoshenko梁非线性动力分析的能量守恒逐步积分算法。采用共旋技术考虑结构的几何非线性,空间离散采用相关插值形式,避免了剪切锁定现象。在时间离散时利用多参数修正方法对等效的节点动力平衡方程进行修正,实现了离散系统在保守荷载作用下的能量守恒。算法具备二阶局部精度,与已有的平均加速度方法和隐式中点方法相比,具有更好的数值稳定性。在二维情形下与Simo方法对比,指出了Simo方法在受保守外弯矩作用时系统能量不守恒。最后,通过三个数值模拟算例验证了算法的性能和能量守恒特性。     

On the a-priori error analysis and convergence of meshless BEM

This paper deals with the a-priori error analysis and convergence of meshless boundary element methods. The paper investigates the convergence of the particular solution method (PSM) in conjunction with the boundary element method (BEM) as well as the method of fundamental solutions (MFS) and the radial basis functions (RBF)—type techniques. A-priori error estimates for meshless BEMs are also provided, and several illustrating numerical experiments are derived.     

A practical examination of the errors arising in the direct collocation boundary element method for acoustic scattering

For many engineers and acousticians, the boundary element method (BEM) provides an invaluable tool in the analysis of complex problems. It is particularly well suited for the examination of acoustical problems within large domains. Unsurprisingly, the widespread application of the BEM continues to produce an academic interest in the methodology. New algorithms and techniques are still being proposed, to extend the functionality of the BEM, and to compute the required numerical tasks with greater accuracy and efficiency. However, for a given global error constraint, the actual computational accuracy that is required from the various numerical procedures is not often discussed. Within this context, this paper presents an investigation into the discretisation and computational errors that arise in the BEM for acoustic scattering. First, accurate routines to compute regular, weakly singular, and nearly weakly singular integral kernels are examined. These are then used to illustrate the effect of the requisite boundary discretisation on the global error. The effects of geometric and impedance singularities are also considered. Subsequently, the actual integration accuracy required to maintain a given global error constraint is established. Several regular and irregular scattering examples are investigated, and empirical parameter guidelines are provided.     

The Scalar Homotopy Method for Solving Non-Linear Obstacle Problem

In this study, the nonlinear obstacle problems, which are also known as the nonlinear free boundary problems, are analyzed by the scalar homotopy method (SHM) and the finite difference method. The one- and two-dimensional nonlinear obstacle problems, formulated as the nonlinear complementarity problems (NCPs), are discretized by the finite difference method and form a system of nonlinear algebraic equations (NAEs) with the aid of Fischer-Burmeister NCP-function. Additionally, the system of NAEs is solved by the SHM, which is globally convergent and can get rid of calculating the inverse of Jacobian matrix. In SHM, by introducing a scalar homotopy function and a fictitious time, the NAEs are transformed to the ordinary differential equations (ODEs), which can be integrated numerically to obtain the solutions of NAEs. Owing to the characteristic of global convergence in SHM, the restart algorithm is adopted to fasten the convergence of numerical integration for ODEs. Several numerical examples are provided to validate the efficiency and consistency of the proposed scheme. Besides, some factors, which might influence on the accuracy of the numerical results, are examined by a series of numerical experiments.     

小周期结构Helmholtz方程的多尺度计算

本文研究小周期结构Helmholtz方程的多尺度计算。我们用各向异性多尺度方法(HMM)求解小周期结构Helmholtz问题。借助于渐近分析技术,在对HMM方法深入分析的基础上,我们给出了精确与HMM方法近似解之间的误差估计,并讨论和分析了利用微结构信息校正HMM逼近解的技巧。最后,我们用数值例了验证了理论结果的正确性。