Enriched goal-oriented error estimation for fracture problems solved by continuum-based shell extended finite element method简

An enriched goal-oriented error estimation method with extended degrees of freedom is developed to estimate the error in the continuum-based shell extended finite element method. It leads to high quality local error bounds in three-dimensional fracture mechanics simulation which involves enrichments to solve the singularity in crack tip. This enriched goal-oriented error estimation gives a chance to evaluate this continuum- based shell extended finite element method simulation. With comparisons of reliability to the stress intensity factor calculation in stretching and bending, the accuracy of the continuum-based shell extended finite element method simulation is evaluated, and the reason of error is discussed.     

改进型扩展比例边界有限元法

结合了扩展有限元法(extended finite elementmethods,XFEM)和比例边界有限元法(scaled boundary finite elementmethods,SBFEM)的主要优点,提出了一种改进型扩展比例边界有限元法(improvedextended scaled boundary finite elementmethods,$i$XSBFEM),为断裂问题模拟提供了一条新的途径.类似XFEM,采用两个正交的水平集函数表征材料内部裂纹面,并基于水平集函数判断单元切割类型;将被裂纹切割的单元作为SBFE的子域处理,采用SBFEM求解单元刚度矩阵,从而避免了XFEM中求解不连续单元刚度矩阵需要进一步进行单元子划分的缺陷;同时,借助XFEM的主要思想,将裂纹与单元边界交点的真实位移作为单元结点的附加自由度考虑,赋予了单元结点附加自由度明确的物理意义,可以直接根据位移求解结果得出裂纹与单元边界交点的位移;对于含有裂尖的单元,选取围绕裂尖单元一圈的若干层单元作为超级单元,并将此超级单元作为SBFE的一个子域求解刚度矩阵,超级单元内部的结点位移可通过SBFE的位移模式求解得到,应力强度因子可基于裂尖处的奇异位移(应力)直接获得,无需借助其他的数值方法.最后,通过若干数值算例验证了建议的$i$XSBFEM的有效性,相比于常规XFEM,$i$XSBFEM的基于位移范数的相对误差收敛性较好;采用$i$XSBFEM通过应力法和位移法直接计算得到的裂尖应力强度因子均与解析解吻合\较好.      

Goal-oriented error estimation applied to direct solution of steady-state analysis with frequency-domain finite element method

Based on the concept of the constitutive relation error along with the residuals of both the origin and the dual problems,a goal-oriented error estimation method with extended degrees of freedom is developed.It leads to the high quality local error bounds in the problem of the direct-solution steady-state dynamic analysis with a frequency-domain finite element,which involves the enrichments with plural variable basis functions.The solution of the steady-state dynamic procedure calculates the harmonic response directly in terms of the physical degrees of freedom in the model,which uses the mass,damping,and stiffness matrices of the system.A three-dimensional finite element example is carried out to illustrate the computational procedures.     

A-posteriori error estimation for second order mechanical systems

One important issue for the simulation of flexible multibody systems is the reduction of the flexible bodies degrees of freedom.As far as safety questions are concerned knowledge about the error introduced by the reduction of the flexible degrees of freedom is helpful and very important.In this work,an a-posteriori error estimator for linear first order systems is extended for error estimation of mechanical second order systems.Due to the special second order structure of mechanical systems,an improvement of the a-posteriori error estimator is achieved.A major advantage of the a-posteriori error estimator is that the estimator is independent of the used reduction technique.Therefore,it can be used for moment-matching based,Gramian matrices based or modal based model reduction techniques.The capability of the proposed technique is demonstrated by the a-posteriori error estimation of a mechanical system,and a sensitivity analysis of the parameters involved in the error estimation process is conducted.     

Adaptive non‐conformal mesh refinement and extended finite element method for viscous flow inside complex moving geometries

Two different techniques to analyze non‐Newtonian viscous flow in complex geometries with internal moving parts and narrow gaps are compared. The first technique is a non‐conforming mesh refinement approach based on the fictitious domain method (FDM), and the second one is the extended finite element method (XFEM). The refinement technique uses one fixed reference mesh, and to impose continuity across non‐conforming regions, constraints using Lagrangian multipliers are used. The size of elements locally in the high shear rate regions is reduced to increase accuracy. FDM is shown to have limitations; therefore, XFEM is applied to decouple the fluid from the internal moving rigid bodies. In XFEM, the discontinuous field variables are captured by using virtual degrees of freedom that serve as enrichment and by applying special integration over the intersected elements. The accuracy of the two methods is demonstrated by direct comparison with results of a boundary‐fitted mesh applied to a two‐dimensional cross section of a twin‐screw extruder. Compared with non‐conforming FDM, XFEM shows a considerable improvement in accuracy around the rigid body, especially in the narrow gap regions. Copyright © 2011 John Wiley & Sons, Ltd.     

扩展有限元裂尖场精度研究

论述了扩展有限元方法和基本原理,研究了单元类型(四边形单元和三角形单元、线性单元和二次单元)、网格密度、J积分区域半径等因素对裂尖局部应力场(应力强度因子)计算精度的影响。研究发现,上述因素对裂尖应力强度因子计算的收敛速度与稳定性影响不大,证实了XFEM可以用较少的节点获得较高的裂尖场精度,并提出了通过固定裂尖附加区半径可以进一步改善XFEM的收敛速度。     

一种XFEM断裂分析的裂尖单元新型改进函数

提出了一种适用于裂尖改进单元的新型改进函数, 基于三角变换的方法, 保留裂纹尖端场的应力奇异性和裂纹上、下表面的位移不连续性, 将常规扩展有限元法裂尖改进单元的4 项改进函数缩减为2 项, 裂尖改进单元的结点由常规的8 个改进自由度减少为4 个. 采用2 个正交的水平集函数表征材料内部裂纹面, 详细阐述了改进单元类型的判别方法, 给出一种改进单元的分区域积分方案. 最后, 若干断裂力学问题经典算例的数值计算结果表明:建议的裂尖改进函数具有较高的数值精度, 该方法是十分有效的.     

基于扩展有限元的应力强度因子的位移外推法

针对平面裂纹问题,阐述了扩展有限元法的单元位移模式、推导了扩展有限元法的控制方程、介绍了特殊单元的数值积分技术.基于最小二乘法,建立了应力强度因子位移外推法的计算公式.利用MATLAB编写计算程序,对平面裂纹问题用扩展有限元法进行了计算.基于扩展有限元法的计算结果,分别利用位移外推法和相互作用积分法,对平面裂纹的应力强度因子进行了计算.计算结果表明,位移外推法比相互作用积分法能更方便和准确地计算平面裂纹的应力强度因子.     

基于p型有限元法和围线积分法计算复合型应力强度因子

传统有限元法由于采用低阶插值计算应力强度因子时,需要划分的网格数较多,收敛速度较慢,得到的应力强度因子精度不足。p型有限元法在网格确定时通过增加插值多项式的阶数来提高计算精度,具有网格划分少、收敛速度快、精度高、自适应能力强等特点。本文采用基于p型有限元法的有限元计算软件StressCheck计算得到应力场和位移场,并由围线积分法导出混合型应力强度因子(SIFs)。通过几个经典算例,分析了围线的选择对计算精度的影响,计算了不同裂纹长度、不同裂纹角度和裂纹在应力集中区域不同位置时的应力强度因子。并将数值结果、理论解与文献中其他数值计算方法所得的部分结果进行了对比分析,结果表明自由度数不大于7000时,导出的应力强度因子相对误差最大不超过1.2%,数值解表现出较高的精度及数值稳定性。     

准脆性材料中强不连续问题的数值方法若干应用及其研究进展

综述了模拟准脆性材料开裂过程的数值计算方法的研究进展和工程应用,比较了表征强不连续问题的显式非连续模型和隐式非连续模型的优缺点.结合混凝土粘结裂纹, 重点讨论了嵌入非连续模型,扩展有限元方法和富集有限元技术等非连续方法的构造特征和本质区别.从各种富集方法的理论完备性考察,以假定发展应变为基础的嵌入非连续方法虽然可以解决混凝土开裂过程中的应力锁死,满足内部边界的静力平衡条件以及反映开裂后的位移不连续问题,但嵌入非连续所采用的富集函数在开裂单元中并不能满足协调条件,使非连续两侧的应变不独立. 其局限性是由于富集自由度在单元的水平上引入,而以单位分解为基础的扩展有限元和富集有限元的富集函数以节点自由度的方式引入,除具有嵌入非连续的优点, 还可以有效消除嵌入非连续引起裂纹两侧应变的相互影响.文中同时指出了网格重构技术,弥散裂纹模型的局限性以及扩展有限元和富集有限元技术在构造方式上的细微差别.对于节点自由度方式引入的富集函数, 其操作困难性在文中也作了说明.      

直接计算应力强度因子的扩展有限元法

系统地给出了直接计算应力强度因子的扩展有限元法。该方法以常规有限元法为基础,利用单位分解法思想,通过在近似位移表达式中增加能够反映裂纹面的不连续函数及反映裂尖局部特性的裂尖渐进位移场函数,间接体现裂纹面的存在,从而无需使裂纹面与有限元网格一致,无需在裂尖布置高密度网格,也不需要后处理就可以直接计算出应力强度因子,并且大大简化了前后处理工作。最后通过两个简单算例验证了该方法的精度,分析了影响计算结果的因素,并与采用J积分计算的应力强度因子作了对比,得出了两种方法计算精度相当的结论。     

孔洞对平面裂纹应力强度因子影响分析的广义参数Williams单元

本文采用圆形奇异区广义参数Williams单元（W单元）建立了中心裂纹与圆孔共存的平面应力模型,奇异区外围利用ABAQUS有限元软件自动网格离散技术与FORTRAN95编程前处理相结合,克服了自主编程中网格离散的局限性．算例分析了圆孔位置和几何参数对I-II混合型裂纹尖端应力强度因子(SIFs)的影响,并与扩展有限元法(XFEM)计算结果进行比较．结果表明：靠近圆孔一侧的裂尖SIFs大于远离圆孔一侧的裂尖SIFs;控制圆孔左边缘到裂纹中心的距离,则两侧裂尖SIFs随圆孔半径的增大而增大;圆孔中心与裂纹中心水平距离越远,圆孔对裂纹扩展的影响越小．同时,基于圆形奇异区的W单元直接计算得到的裂尖SIFs与扩展有限元法得到的解吻合较好,证明了W单元对奇异区离散形状不敏感,且具有高效率和高精度．     

Thermoelastic analysis of multiple defects with the extended finite element method

In this paper, the extended finite element method (XFEM) is adopted to analyze the interaction between a sin-gle macroscopic inclusion and a single macroscopic crack as well as that between multiple macroscopic or micro-scopic defects under thermal/mechanical load. The effects of different shapes of multiple inclusions on the material thermomechanical response are investigated, and the level set method is coupled with XFEM to analyze the interaction of multiple defects. Further, the discretized extended finite element approximations in relation to thermoelastic prob-lems of multiple defects under displacement or temperature field are given. Also, the interfaces of cracks or materials are represented by level set functions, which allow the mesh assignment not to conform to crack or material interfaces. Moreover, stress intensity factors of cracks are obtained by the interaction integral method or the M-integral method, and the stress/strain/stiffness fields are simulated in the case of multiple cracks or multiple inclusions. Finally, some numer-ical examples are provided to demonstrate the accuracy of our proposed method.     

The intrinsic XFEM for two‐fluid flows

In two‐fluid flows, jumps and/or kinks along the interfaces are present in the resulting velocity and pressure fields. Standard methods require mesh manipulations with the aim that either element edges align with the interfaces or that the mesh is sufficiently refined near the interfaces. In contrast, enriched methods, such as the extended finite element method (XFEM), enable the representation of arbitrary jumps and kinks inside elements. Thereby, optimal convergence can be achieved for two‐fluid flows with meshes that remain fixed throughout the simulation. In the intrinsic XFEM, in contrast to other enriched methods, no more unknowns are present in the approximation than in a standard finite element approximation. In this work, the intrinsic XFEM is employed for the simulation of incompressible two‐fluid flows. Numerical results are shown for a number of test cases and prove the success of the method. Copyright © 2008 John Wiley & Sons, Ltd.     

Evolving crack patterns in thin films with the extended finite element method

This paper develops the extended finite element method (XFEM) to evolve patterns of multiple cracks, in a brittle thin film bonded to an elastic substrate, with a relatively coarse mesh, and without remeshing during evolution. A shear lag model describes the deformation in three dimensions with approximate field equations in two-dimensions. The film is susceptible to subcritical cracking, obeying a kinetic law that relates the velocity of each crack to its energy release rate. At a given time, the XFEM solves the field equations and calculates the energy release rate of every crack. For a small time step, each crack is extended in the direction of maximal hoop stress, and by a length set by the kinetic law. To confirm the accuracy of the XFEM, we compare our simulation to the exiting solutions for several simple crack patterns, such as a single crack and a set of parallel cracks. We then simulate the evolution of multiple cracks, initially in a small region of the film but of different lengths, showing curved crack propagation and crack tip shielding. Starting with multiple small cracks throughout the film, the XFEM can generate the well-known mud crack pattern.     

动载下裂纹应力强度因子计算的改进型扩展有限元法

相较于常规扩展有限元法(extended finite element method, XFEM), 改进型扩展有限元法(improved XFEM) 解决了现有方法线性相关与总体刚度矩阵高度病态问题, 在数量级上提升了总体方程的求解效率, 克服了现有方法在动力学问题中的能量正确传递、动态应力强度因子数值震荡、精度低下问题. 本文基于改进型XFEM, 采用Newmark 隐式时间积分算法, 重点研究了动载荷作用下扩展裂纹尖端应力强度因子的求解方法, 与静力学方法相比, 增加了裂纹扩展速度项与惯性项的贡献. 通过数值算例研究了网格单元尺寸、质量矩阵、时间步长、裂尖加强区域、惯性项、扩展速度项及相互作用积分区域J-domain的网格与单元尺寸对动态应力强度因子求解精度的影响, 验证了改进型XFEM计算动态裂纹应力强度因子方法的有效性. 针对文献中具有挑战性的 "I 型半无限长裂纹先稳定后扩展"问题, 改进型XFEM给出目前为止精度最好的动态应力强度因子数值解.      

广义扩展有限元法及其在裂纹扩展分析中的应用

结合广义有限元法(GFEM)和扩展有限元法(XFEM)的特点,提出了一种新的数值方法——广义扩展有限元法(GXFEM)。阐述了广义扩展有限元法的基本原理,对相关公式进行推导,探讨数值实施中需注意的重要问题,给出利用广义扩展有限元法进行断裂分析时应力强度因子的计算方法,编写了广义扩展有限元法程序。通过算例进行了应力强度因子的计算,模拟了结构裂纹的扩展过程。算例结果表明,利用广义扩展有限元法计算裂纹扩展问题,不需要进行过密的网格划分,且网格在裂纹扩展后无需重新剖分,具有相当高的计算精度。     

An extended finite element method for modeling near-interfacial crack propagation in a layered structure

The extended finite element method (XFEM) is applied for the simulation of near-interfacial crack propagation in a metal–ceramic layered structure. An experimental evidence indicates that, in a ceramic–metal–ceramic sandwich structure, a near-interfacial crack in the ceramic layer can be drawn to or deflect away from the metal layer depending on the difference in elastic properties across the interface. To model near-interfacial fracture, only the Heaviside functions are used for the XFEM, and the vector level set method is employed for efficient evaluation of the enrichment functions. The crack propagation paths predicted by the XFEM simulation are found to be consistent with the experimental observation.     

THE EVALUATION OF STRESS INTENSITY FACTORS OF PLANE CRACK FOR ORTHOTROPIC PLATE WITH EQUAL PARAMETER BY F2LFEM

In this paper, the evaluation of stress intensity factor of plane crack problems for orthotropic plate of equal-parameter is investigated using a fractal two-level finite element method (F2LFEM). The general solution of an orthotropic crack problem is obtained by assimilating the problem with isotropic crack problem, and is employed as the global interpolation function in F2LFEM. In the neighborhood of crack tip of the crack plate, the fractal geometry concept is introduced to achieve the similar meshes having similarity ratio less than one and generate an infinitesimal mesh so that the relationship between the stiffness matrices of two adjacent layers is equal. A large number of degrees of freedom around the crack tip are transformed to a small set of generalized coordinates. Numerical examples show that this method is efficient and accurate in evaluating the stress intensity factor (SIF).     

A domain-independent interaction integral for fracture analysis of nonhomogeneous piezoelectric materials

Piezoelectric materials and structures contain more or less electromechanical interfaces in engineering applications. It is difficult to obtain the fracture parameters efficiently of the piezoelectric materials with complex interfaces. This paper presents a domain-independent interaction integral for material nonhomogeneity and discontinuity which can be used for solving the stress intensity factors (SIFs) and the electric displacement intensity factor (EDIF) of piezoelectric materials with complex interfaces efficiently. The interaction integral is based on the J-integral by superimposition of two admissible states and the present formulation does not involve any derivatives of mechanical and electric properties. Moreover, it is proved that the interface in the integral domain does not affect the value of the interaction integral and thus, the present method does not require electromechanical parameters of piezoelectric materials to be continuous. The interaction integral method combined with the extended finite element method (XFEM) is used to investigate the influences of material continuity on the SIF and the EDIF and the results show that the material parameters and their first-order derivatives affect both the SIF and the EDIF greatly, while the higher-order derivatives affect both of them slightly.