资源描述
附录1：外文翻译CAE的技术种类有很多，其中包括有限元法,边界元法,有限差法等。每一种方法各有其应用的领域，而其中有限元法应用的领域越来越广，现已应用于结构力学、结构动力学、热力学、流体力学、电路学、电磁学等。ANSYS软件是融结构、流体、电场、磁场、声场分析于一体的大型通用有限元分析软件。由世界上最大的有限元分析软件公司之一的美国ANSYS开发，它能与多数CAD软件接口，实现数据的共享和交换，如Pro/Engineer, NASTRAN, Alogor, IDEAS, AutoCAD等，是现代产品设计中的高级CAE工具之一。 ANSYS有限元软件包是一个多用途的有限元法计算机设计程序，可以用来求解结构、流体、电力、电磁场及碰撞等问题。因此它可应用于以下工业领域：航空航天、汽车工业、生物医学、桥梁、建筑、电子产品、重型机械、微机电系统、运动器械等。有限元分析（FEA，Finite Element Analysis）的基本概念是用较简单的问题代替复杂问题后再求解。它将求解域看成是由许多称为有限元的小的互连子域组成，对每一单元假定一个合适的(较简单的）近似解，然后推导求解这个域总的满足条件(如结构的平衡条件），从而得到问题的解。这个解不是准确解，而是近似解，因为实际问题被较简单的问题所代替。由于大多数实际问题难以得到准确解，而有限元不仅计算精度高，而且能适应各种复杂形状，因而成为行之有效的工程分析手段。 有限元是那些集合在一起能够表示实际连续域的离散单元。有限元的概念早在几个世纪前就已产生并得到了应用，例如用多边形（有限个直线单元）逼近圆来求得圆的周长，但作为一种方法而被提出，则是最近的事。有限元法最初被称为矩阵近似方法，应用于航空器的结构强度计算，并由于其方便性、实用性和有效性而引起从事力学研究的科学家的浓厚兴趣。经过短短数十年的努力，随着计算机技术的快速发展和普及，有限元方法迅速从结构工程强度分析计算扩展到几乎所有的科学技术领域，成为一种丰富多彩、应用广泛并且实用高效的数值分析方法。 有限元方法与其他求解边值问题近似方法的根本区别在于它的近似性仅限于相对小的子域中。20世纪60年代初首次提出结构力学计算有限元概念的克拉夫（Clough）教授形象地将其描绘为：“有限元法=Rayleigh Ritz法分片函数”，即有限元法是Rayleigh Ritz法的一种局部化情况。不同于求解（往往是困难的）满足整个定义域边界条件的允许函数的Rayleigh Ritz法，有限元法将函数定义在简单几何形状（如二维问题中的三角形或任意四边形）的单元域上（分片函数），且不考虑整个定义域的复杂边界条件，这是有限元法优于其他近似方法的原因之一。 对于不同物理性质和数学模型的问题，有限元求解法的基本步骤是相同的，只是具体公式推导和运算求解不同。有限元求解问题的基本步骤通常为： 第一步：问题及求解域定义：根据实际问题近似确定求解域的物理性质和几何区域。 第二步：求解域离散化：将求解域近似为具有不同有限大小和形状且彼此相连的有限个单元组成的离散域，习惯上称为有限元网络划分。显然单元越小（网络越细）则离散域的近似程度越好，计算结果也越精确，但计算量及误差都将增大，因此求解域的离散化是有限元法的核心技术之一。 第三步：确定状态变量及控制方法：一个具体的物理问题通常可以用一组包含问题状态变量边界条件的微分方程式表示，为适合有限元求解，通常将微分方程化为等价的泛函形式。 第四步：单元推导：对单元构造一个适合的近似解，即推导有限单元的列式，其中包括选择合理的单元坐标系，建立单元试函数，以某种方法给出单元各状态变量的离散关系，从而形成单元矩阵（结构力学中称刚度阵或柔度阵）。为保证问题求解的收敛性，单元推导有许多原则要遵循。 对工程应用而言，重要的是应注意每一种单元的解题性能与约束。例如，单元形状应以规则为好，畸形时不仅精度低，而且有缺秩的危险，将导致无法求解。 第五步：总装求解：将单元总装形成离散域的总矩阵方程（联合方程组），反映对近似求解域的离散域的要求，即单元函数的连续性要满足一定的连续条件。总装是在相邻单元结点进行，状态变量及其导数（可能的话）连续性建立在结点处。第六步：联立方程组求解和结果解释：有限元法最终导致联立方程组。联立方程组的求解可用直接法、选代法和随机法。求解结果是单元结点处状态变量的近似值。对于计算结果的质量，将通过与设计准则提供的允许值比较来评价并确定是否需要重复计算。 简言之，有限元分析可分成三个阶段，前处理、处理和后处理。前处理是建立有限元模型，完成单元网格划分；后处理则是采集处理分析结果，使用户能简便提取信息，了解计算结果。 实践中，有限元分析法通常由三个主要步骤组成： 1、预处理：用户需建立物体待分析部分的模型，在此模型中，该部分的几何形状被分割成若干个离散的子区域或称为“单元”。各单元在一些称为“结点”的离散点上相互连接。这些结点中有的有固定的位移，而其余的有给定的载荷。准备这样的模型可能极其耗费时间，所以商用程序之间的相互竞争就在于：如何用最友好的图形化界面的“预处理模块”，来帮助用户完成这项繁琐乏味的工作。有些预处理模块作为计算机化的画图和设计过程的组成部分，可在先前存在的CAD文件中覆盖网格，因而可以方便地完成有限元分析。 2、分析：把预处理模块准备好的数据输入到有限元程序中，从而构成并求解用线性或非线性代数方程表示的系统。Kij*Uj=Fi式中，u 和 f 分别为各结点的位移和作用的外力。矩阵 K 的形式取决于求解问题的类型，本模块将概述桁架与线弹性体应力分析的方法。商用程序可能带有非常大的单元库，不同类型的单元适用于范围广泛的各类问题。有限元法的主要优点之一就是：许多不同类型的问题都可用相同的程序来处理，区别仅在于从单元库中指定适合于不同问题的单元类型。 3、后处理：在有限元分析的早期，用户需仔细地研读程序运算后产生的大量数字，即列出的模型内各离散位置处的位移和应力。这种方法容易漏掉重要的趋向与热点，而最新的程序则利用图形显示来帮助用户直接观察运算结果。典型的后处理模块能显示遍布于模型上的彩色等应力线图，以表示不同的应力水平，显示的整个应力场的图像类似于光弹性法或云纹法的实验结果。附录2：外文原文出处：Finite element overviewThere are many types of CAE technology, including the finite element method, boundary element method, finite difference method. Each method has its own application areas, of which the application of finite element method more and more areas, has been used in structural mechanics, structural dynamics, thermodynamics, fluid mechanics, circuit theory, electromagnetism and so on. ANSYS software is the financial structure, fluid, electric field, magnetic field, acoustic field analysis in one large-scale finite element analysis software. By the worlds largest finite element analysis software ANSYS, one of the United States developed it with most CAD software interface for data sharing and exchange, such as Pro / Engineer, NASTRAN, Alogor, I-DEAS, AutoCAD, are modern Advanced CAE product design tools. ANSYS finite element package is a multi-purpose finite element method for computer design program, can be used to solve the structure, fluid, electricity, electromagnetic fields and collision issues. So it can be applied to the following industries: aerospace, automotive, biomedical, bridges, construction, electronics, heavy machinery, micro-electromechanical systems, sports equipment, etc. Finite Element Analysis (FEA， Finite Element Analysis) of the basic concept is to re-place the relatively simple problem to solve complex problems later. As it will solve the do-main is composed of many small-called finite element subdomain interconnection compone-nts，assuming that each unit of an appropriate (relatively simple) approximate solution， and then derived the general solution of the domain satisfy the conditions (such as balanced con-ditions)， thus the solution of the problem. This solution is not exact solutions，but appro-ximate solution， since the actual problem is relatively simple to replace the problem. Since most practical problems it is difficult to be accurate solution， while finite element is not only high accuracy but also to adapt to a variety of complex shapes， thereby becoming an effective means of engineering analysis. FEM together those who are able to express the actual domain for the discrete element. The concept of the finite element as early as several centuries ago and have been applied， for example， polygon (a finite number of straight-line unit) to get close to circle the cir-cumference of a circle， but as a way to be made， it is the most recent matter. Finite ele-ment method was originally known as the matrix approximation method， the structural strength of aircraft used in the calculation， and because of its convenience， practicality and effectiveness arising from research scientists to engage in mechanical interest. Through the efforts of just a few decades， with the rapid development of computer technology and the popularity of the finite element method in structural engineering from the intensity of the rapid analysis extended to almost all areas of science and technology， become a rich and colorful， practical and efficient application of a wide range of numerical analysis. Finite element method with other methods of solving the boundary value problem simil-ar to the fundamental difference is that the approximation of it is limited to relatively small sub-domain. 60 In the early 20th century structure was first proposed the concept of the finite element calculation of Clough (Clough)， Professor vividly describes as: The finite element method + = Rayleigh Ritz method piecewise function， that is， the finite element method is the Rayleigh Ritz method a localized situation. Different from the solution of (often difficult) to satisfy the boundary conditions of the definition of domain function to allow the Rayleigh Ritz method， finite element method will be defined in a simple function of geometry (such as two-dimensional problem of arbitrary quadrilateral or triangle) on the unit domain ( piecewise function)， the definition does not consider the whole domain of the complex boundary conditions， this is the finite element method is superior to other similar methods of one of the reasons why. Different physical properties and mathematical models of the problem， finite element method to solve the basic steps are the same， only the specific formula to solve a different derivation and computation. Finite Element Analysis of the basic steps are as follows: The first step: the definition of the problem and solution domain: In accordance with the actual problem solving domain approximation to determine the physical properties and geometry of the region. The second step: Solving domain discretization: The approximate solution of the domain with different size and shape of a limited and linked to each other unit， composed of a fin-ite number of discrete domains， the habit of division as the finite element network. Obvio-usly the smaller the unit (the finer t he network) is similar to the level of discrete domain， the better，the more accurate results， but the calculation of the volume and error will be larger，so to solve the discrete domain is the finite element method，one of the core tech-nology. The third step: to determine the state variables and control method: a specific physical problem can usually be handled by a group of state variables include the issue of boundary conditions that the differential equations for the finite element for solving differential equa-tions are usually translated into the functional equivalent forms of . Step four: unit derived: on the unit to construct a suitable approximate solution， that is derived out of the finite element type， including a reasonable choice of coordinate system units， the establishment of unit test function， to one way or another unit of the state va-riables given the discrete relations to form the unit matrix (the structure of said mechani-cal stiffness or flexibility matrix array). In order to ensure the convergence of problem solving， there are many principles de-rived units to follow. In terms of engineering applications， it is important to pay atten-tion to each unit of problem-solving performance and constraints. For example， the unit should be based on the rules for shape， and deformed not only low-precision， but also the risk of missing rank， will result in failure to solve. Step five: Solution assembly: assembly to form a discrete unit of the total domain matrix equation (Joint equations)， reflecting the approximate solution of the discrete domain the request domain， that is， the continuity of function modules to meet the conditions for cer-tain. Assembly unit in the adjacent node， the state variables and their derivatives (if possib-le) to establish continuity in the junction point. Sixth step: solving simultaneous equations and the results of the interpretation: the finite element method eventually lead to simultaneous equations. Simultaneous equations can be used to solve the direct method， the election law and the random generation method. Solv-ing a result， the state Department unit node approximation variables. The results for the quality and design guidelines will be provided to allow values to evaluate and determine the need for double-counting. In short， the finite element analysis can be divided into three stages， pre-treatment， processing and post-processing. Pre-processing finite element model is built to complete the unit mesh; post-processing is the acquisition and processing the results of the analysis， a-lows users to extract information easy to understand results. In practice， the finite element method is usually composed of three main steps: 1， pre-processing: the user object to be analyzed to establish part of the model， in this model， the geometry of the part being cut into several discrete sub-region - otherwise known as modules. In some of the modules referred to as nodes of the discrete points connected with each other. Some of these nodes are fixed displacement， while the remaining loads are given. Prepare such a model could be extremely time-consuming process is why the commercial competition between the lies: how to use the most friendly graphical inter-face of the pre-processing module， to help users complete the tedious work of boring. Some pre-processing module as a computerized drawing and an integral part of the de-ign process， can be pre-existing CAD file grid coverage， which can be easily completed by Finite Element Analysis. 2， Analysis: the pre-processing module prepared data into finite element program,and thus constitutes a solution of linear or nonlinear system of algebraic equations that Kij * Uj = Fi Where u and f，respectively，for each node of the displacement and the role of external forces. Matrix form of K depend on the type of problem solving， the module will outline the truss with the linear elastic stress analysis. Business procedures may carry a very large library， the different types of unit s applicable to a wide range of various problems. Finite element method is one of the main advantages of: Many different types of problems are available to deal with the same procedure， the difference is only specified from the cell library for the problem in different cell types. 3， post-processing: In the early finite element analysis， users need to carefully study the procedures for computing a large number of figures after， that is， the model set out in the discrete position of the displacement and stress. This method is easy to miss important trends and hot spots， and the latest graphics processing to be use to help the user com-puting the results of direct observation. Typical post-processing module can display the model across the color line graph of stress for different stress levels， indicating the entire stress field is similar to the images or Photoelasticity moire results.
展开阅读全文