Use of Tikhonov’s regularization method for solving identification problem for elastic systems

We present a method for solving nonlinear inverse problems, which also include identification problems for elastic systems. The problems whose initial data contain an error are usually solved by regularization methods [1–5]. In the present paper, we give preference to Tikhonov’s regularization method, which has been widely used in the recent years in practice to increase the stability of computational algorithms for solving problems in various areas of mechanics [6–9].     

A momentum coupling method for the unsteady incompressible Navier–Stokes equations on the staggered grid

A new numerical method is developed to efficiently solve the unsteady incompressible Navier–Stokes equations with second-order accuracy in time and space. In contrast to the SIMPLE algorithms, the present formulation directly solves the discrete x- and y-momentum equations in a coupled form. It is found that the present implicit formulation retrieves some cross convection terms overlooked by the conventional iterative methods, which contribute to accuracy and fast convergence. The finite volume method is applied on the fully staggered grid to solve the vector-form momentum equations. The preconditioned conjugate gradient squared method (PCGS) has proved very efficient in solving the associate linearized large, sparse block-matrix system. Comparison with the SIMPLE algorithm has indicated that the present momentum coupling method is fast and robust in solving unsteady as well as steady viscous flow problems. © 1998 John Wiley & Sons, Ltd.     

Calculation of the scattering of elastic waves from a penny-shaped crack by the method of optimal truncation

The method of optimal truncation (MOOT) [1, 2], a least-squares boundary-residual method [3] for solving scattering problems, is applied to the plane circular crack. An equatorially cloven spherical inclusion is used to model the crack. Numerical advantages of this model are discussed and demonstrated. Results are given for cross sections for longitudinal waves incident on the crack at arbitrary angles. Both clear cracks and fluid-filled cracks are considered. A refinement of the method which would allow accurate calculation of dynamic stress-intensity factors is developed.     

Supersonic flow over elongated blunt bodies with a cross section of elliptical shape

Up to now computational algorithms have been developed for, and systematic studies have been made of, supersonic flow over axisymmetric bodies both by a stream of ideal gas and by an air stream with equilibrium and nonequilibrium physicochemical transformations [1–6]. Conical flows around bodies having cross sections of different shapes and in a wide range of angles of attack have been studied in detail [7–11]. With the further development of numerical methods the next problem has become the analysis of supersonic flow over blunt bodies of large elongation having cross sections of sufficiently arbitrary shape. The effects of essentially three-dimensional flow (without planes of symmetry) over bodies whose cross sections represent ellipses with a constant or variable ratio of axes along the length of the body are discussed in the present paper.Translated from Izvestiya Akademii Nauk SSSR, Mekhanika Zhidkosti i Gaza, No. 6. pp. 155–159, November–December, 1976.     

Two‐dimensional problems of beam forming under conditions of creep

Direct and inverse problems of forming of longlength profiles with double curvature and a given angle of twisting under conditions of creep are considered. A finitedifference scheme for the numerical solution is proposed. Examples of solving problems with different types of external actions for a profile with a rectangular cross section are given. Experimental and numerical data are compared for twisting of beams with square and circular cross sections in the regime of creep at temperatures of 725 and 740° C for St. 45 steel.     

Scattering of guided waves by defective adhesive bonds in multilayer anisotropic plates

We introduce an analytic-numerical method to simulate the interaction between ultrasonic guided waves and defective adhesive bonds in multilayer structures. We replace the thin adhesive layer by equivalent continuous distribution of normal and transversal springs and locally reduce the corresponding spring constants to model localized defects. We then formulate the resulting scattering problem as a linear least squares problem and solve it accordingly. The developed formulation is easy to implement and equally well suited to treating anisotropic as well as isotropic constituent layers. We illustrate the application of the proposed method through the simulation of ultrasound inspection of a three-layer isotropic plate and a sixteen-layer anisotropic plate used in the aeronautical industry, indicating in the simulations the guided modes that will most strongly interact with interfacial defects and computing the scattering resulting from the interaction of these modes with localized defects. We believe that the proposed method will serve both to aid in the design of interface inspections and as a basis for solving inverse scattering problems.     

A Finite Element Analysis of a Viscoplastic Plate Subjected to Rapid Heating

Abstract

This paper models the response of a thin metallic plate that is subjected to a rapid heat input. In order to accurately model plate response, both the dynamic mechanical and transient heat transfer problems must be solved. The solution is complicated by nonlinearities due to radiation boundary conditions and material inelasticity. Furthermore, the viscoplastic constitutive equations that model the mechanical material behavior are numerically stiff. Nonlinear finite element algorithms are developed for both heat transfer and mechanical analyses. The algorithms are both stable and efficient for solving the problems considered herein. Example problems presented in the paper demonstrate the importance of including material nonlinearity in the model     

The minimal error method for the Cauchy problem in linear elasticity. Numerical implementation for two-dimensional homogeneous isotropic linear elasticity

In this paper, yet another iterative procedure, namely the minimal error method (MEM), for solving stably the Cauchy problem in linear elasticity is introduced and investigated. Furthermore, this method is compared with another two iterative algorithms, i.e. the conjugate gradient (CGM) and Landweber–Fridman methods (LFM), previously proposed by Marin et al. [Marin, L., Háo, D.N., Lesnic, D., 2002b. Conjugate gradient-boundary element method for the Cauchy problem in elasticity. Quarterly Journal of Mechanics and Applied Mathematics 55, 227–247] and Marin and Lesnic [Marin, L., Lesnic, D., 2005. Boundary element-Landweber method for the Cauchy problem in linear elasticity. IMA Journal of Applied Mathematics 18, 817–825], respectively, in the case of two-dimensional homogeneous isotropic linear elasticity. The inverse problem analysed in this paper is regularized by providing an efficient stopping criterion that ceases the iterative process in order to retrieve stable numerical solutions. The numerical implementation of the aforementioned iterative algorithms is realized by employing the boundary element method (BEM) for two-dimensional homogeneous isotropic linear elastic materials.     

小波方法及其非线性力学问题应用分析

小波分析是近几十年来发展起来的重要数学分支,被誉为“数学显微镜”,其独具的多分辨分析和大量可供选择的,可兼具正交性、紧支性、对称性、低通滤波、线性相位及插值性等优良数学品质的小波基函数为强非线性微分方程的数值求解带来了新的契机。自上世纪90年代以来,诸如小波伽辽金法、小波配点法、小波有限单元法和小波边界单元法等数值方法被先后构建出来并成功应用于各类力学问题的定量研究之中。本文从小波提出的历史背景及作为其理论基础的多分辨分析出发,对现有基于小波理论的各类数值方法进行梳理,总结各自的优点、缺点和下一步可能的发展方向,为未来基于小波理论的定量分析方法的发展及其在复杂非线性力学问题中的应用研究提供参考。     

EXAMPLES ARE GIVEN TO ILLUSTRATE THE UNIQUENESS LAW OF SOLUTIONS FOR PLANE PROBLEMS OF ELASTICITY

解的唯一性定理是用逆解法或半逆解法求解弹性力学问题的理论依据,在此用应力函数法、应力法、应力和函数法求解弹性力学平面问题,让学生切实、深入地理解解的唯一性定理的内在含义,丰富和扩大弹性力学的解题方法和应用范围。     

Development of a family of explicit algorithms for structural dynamics with unconditional stability

A new family of explicit integration algorithms is developed based on discrete control theory for solving the dynamic equations of motion. The proposed algorithms are explicit for both displacement and velocity and require no factorisation of the damping matrix and the stiffness matrix. Therefore, for a system with nonlinear damping and stiffness, the proposed algorithms are more efficient than the common explicit algorithms that provide only explicit displacement. Accuracy and stability properties of the proposed algorithms are analysed theoretically and verified numerically. Certain subfamilies are found to be unconditionally stable for any system state (linear elastic, stiffness softening or stiffness hardening) that may occur in earthquake engineering of a practical structure. With dual explicit expression and excellent stability property, the proposed family of algorithms can potentially solve complicated nonlinear dynamic problems.     

Development of an X-ray computed tomography (CT) system with sparse sources: application to three-phase pipe flow visualization

This paper describes the application of a two-beam X-ray computed tomography (CT) system to multiphase (gas–oil–water) flow measurement. Two high-voltage (160 keV) X-ray sources are used to penetrate a 4-in. (101.6 mm ID) pipeline. A rotating filter wheel mechanism is employed to alternately “harden” and “soften” the X-ray spectra to provide discrimination between the three phases. Because this system offers only two projections, conventional back-projection algorithms are ineffective and thus a new reconstruction technique has been developed. A matrix equation is formed, to which additional “smoothing equations” are added to compensate for the lack of projection data. The tomographic result is obtained by computing an inverse matrix. This is a one-off computation and the inverse is stored for repeated use; reconstructed images from synthesized data demonstrate the effectiveness of this technique. Three-phase tomographic images of a horizontal slug flow are presented, which clearly show the mixing of oil and water layers within the slug body. The relevance of this work to the offshore oil and gas industry is summarized.     

无网格法研究进展及其应用

从加权残量法的角度出发，系统地总结了现有各种无网格法的基本格式，阐明了无网格法的特点，论述了无网格法的研究进展，给出了无网格法在碰撞、动态裂纹扩展、金属加工成型、流体力学以及其它领域中的应用。     

Cross-Properties Relations in 3D Percolation Networks: I. Network Characteristic Length Determination

This work developed efficient methods to determine the hydraulic characteristic length of percolation networks. The proposed methods are based on the determination of the percolation threshold using the Hoshen–Kopelman algorithm extended to 3D networks. The methods were tested with simple cubic site-bond network models with and without spatial correlations. It was shown that the proposed methods outperformed the algorithms based on the simulation of the porosimetry curve.*Author for correspondence: Fax +55-21-2562-8300; e-mail: paulo@peq.coppe.ufrj.br     

正交各向异性弹塑性摩擦接触问题的数值求解

采用正交各向异性摩擦定律对三维弹塑性摩擦接触问题进行分析,基于参变量变分原理,经过有限元离散,将问题化为线性互补问题模型,之后给出一个求解互补问题的非内点光滑化算法．对三维接触问题,滑动方向的确定一直是个难点,为此,该文采用作者提出的组合规划法和迭代法对各向异性摩擦本构模型进行分析,数值结果说明了模型与算法的正确性。     

极端变形问题的物质点法研究进展

物质点法采用了拉格朗日和欧拉双重描述,将物体离散为一组在固定于空间的网格(欧拉描述)中运动的质点(拉格朗日描述),有效地综合了拉格朗日法和欧拉法的优点,是分析冲击爆炸等极端变形问题的一种有效方法. 本文系统总结了本课题组近年来针对冲击、爆炸和流固耦合等极端变形问题在物质点法算法方面的研究进展,并简要介绍了开发的三维显式并行物质点法数值仿真软件MPM3D及其在超高速碰撞、侵彻、爆炸、边坡失效、金属切削和流固耦合等问题中的应用.     

超高速碰撞问题的三维物质点法

简要介绍了物质点法（material point method）的离散原理,通过引入Johnson-Cook材料模型和Mie-Gruneisen状态方程,将其用于超高速碰撞问题的分析中,并编制了相应三维物质点法程序MPM3D。该方法避免了拉格朗日格式因网格畸变产生的数值困难,也克服了欧拉格式材料界面跟踪问题以及因非线性对流扩散项而引起的数值困难。利用该程序对Taylor杆高速碰撞问题和空间碎片防护超高速碰撞问题进行了数值模拟,所得数值结果与实验结果基本吻合,验证了程序的正确性,说明了物质点法在分析超高速碰撞问题时相对于有限元法的优势。     

One method of profiling short plane nozzles

One of the possible methods is considered for profiling short plane nozzles for aerodynamic tubes. The nozzle has a straight sonic line, which allows the subsonic and supersonic sections to be constructed separately. The problem is solved numerically in the plane of a hodograph. In the subsonic region, Dirichlet's problem is formulated for Chaplygin's equation in a rectangle, one side of which is the sonic line. At the present time, two approaches have been defined in papers on calculations of a Laval nozzle, associated with the solution of the so-called “direct” and “inverse” problems (one has in mind a study of the flow in the interconnected region of sub- and supersonic flow). The direct problem determines the flow field in the case of a previously specified contour of the channel wall, the shape of which from technical considerations is obtained with certain geometry conditions. The direct problem can be applied in the construction of the Laval nozzle, if the contour of the inlet section of the channel (generally speaking, quite arbitrary) is chosen so successfully that neither shock compressions nor breakaway zones result in the flow. Although a strictly mathematical theory of the direct problem of the Laval nozzle is only being developed at present, there are still very effective numerical methods for its solution [1, 2]. In the inverse problem (which, by definition, is a problem of profiling), the contour of the nozzle is found with respect to a specified velocity distribution on the axis of symmetry. It is assumed that this quite arbitrary dependence can be selected from the condition of the absence of breakaway zones and shock compressions in the nozzle. By its formulation, the inverse problem is Cauchy's problem which, as is well-known, is incorrect in the classical sense in the ellipticity region — the subsonic section of the nozzle. At present, there are also efficient methods of solving the inverse nozzle problem [3], by interpreting it as an arbitrarily correct problem. Difficulties can arise in the inverse problem, in the provision of short (and, consequently, steep) nozzles because of the sharp increase of the error in the calculation. Together with the stated problems, a procedure can be evolved which is associated with the solution of the correctly posed problem for Chaplygin's equation in the plane of the hodograph. This approach is convenient in that it succeeds a priori in fulfilling the important condition of monotonicity of the velocity at the wall, ensuring (in the absence of shock compressions) nonseparability of the streamline flow at any Reynold's numbers.     

A limited survey of various conjugate gradient methods for solving complex matrix equations arising in electromagnetic wave interactions

There exists a wide variety of conjugate gradient algorithms for solving the complex matrix equations arising in electromagnetic wave interactions. The objective of this paper is to describe several of these algorithms which we have found to be computationally efficient. Both the strong and weak points of each version of the conjugate gradient method are presented. The principal conclusion is that at present, there is no single algorithm that provides an efficient solution for all types of problems. The computer programs corresponding to the algorithms discussed in the paper are provided in the Appendix.     

A non-periodic spectral method with application to nonlinear water waves

Spectral methods are very efficient and powerful for solving periodic problems. A new spectral method is developed for problems with no spatial periodicity, and demonstrated for water waves. The method splits the potential into the sum of a prescribed non-periodic component and an unknown periodic component. Computed results are compared with experiments by Shemer et al (1998).