膜结构极小曲面找形的一种自适应有限元分析

找形分析是膜结构设计中的关键环节,但在数学上,膜结构的极小曲面找形分析是一个高度非线性问题,一般无法求得其解析解,因此数值方法成为重要工具。近年来,基于单元能量投影法（EEP法）的一维非线性有限元的自适应分析已经取得成功,基于EEP法的二维线性有限元自适应分析也被证实是有效、可靠的。在此基础上,该文提出一种基于EEP法的二维非线性有限元自适应方法,并成功将之应用于膜结构的找形分析。其主要思想是,通过将非线性问题用Newton法线性化,引入现有的二维线性问题的自适应求解技术,进而实现二维有限元自适应分析技术从线性到非线性的跨越,将非线性有限元的自适应分析求解从一维问题拓展到二维问题。该方法兼顾求解的精度和效率,对网格自适应地进行调整,最终得到优化的网格,其解答可按最大模度量逐点满足用户设定的误差限。该文综述介绍了这一进展,并给出数值算例用以表明该方法的可行性和可靠性。     

二维变分不等式问题的自适应有限元分析

有限元后处理超收敛计算的EEP(单元能量投影)法以及基于该法的自适应有限元分析已在一维变分不等式问题的求解中取得显著成功。以此为基础,该文对二维变分不等式问题成功地实现了自适应有限元分析。该文提出二维区域二分法和二维C 检验技术,有效地提升了松弛迭代的收敛速率,进而应用EEP 超收敛公式计算超收敛解答,用其检验误差并指导网格细分。该文给出的典型数值算例表明该文算法高效、稳定、精确,解答可逐点以最大模度量满足精度要求,堪称为数值精确解。     

二维自由振动的有限元线法自适应分析新进展

有限元线法(FEMOL)是一种优良的半解析、半离散方法,可将其比拟为广义一维问题,进而将一维有限元中单元能量投影(EEP)法及相应的自适应求解技术引入,使FEMOL由半解析方法变为完全解析、数值精确的方法。在对二维线性问题成功地实现了自适应FEMOL分析的基础上,该文进一步报道FEMOL自适应方法在二维自由振动问题中的成功应用和最新进展。该文简要介绍了FEMOL自适应分析二维振动问题的求解策略和技术,整套方法思路清晰、算法严谨、高效可靠,可以得到满足精度要求的自振频率和按最大模度量满足用户事先给定误差限的振型,均为数值精确解。该文给出的数值算例表明所提出的算法具有高效、稳定、通用、可靠的优良特性。     

基于EEP超收敛解的自适应有限元法特性分析

基于EEP （单元能量投影）超收敛计算的自适应有限元法,已对一系列问题取得成功,但其自适应特性尚缺乏相关研究。该文以二阶常微分方程为模型问题,同时考察基于EEP和SPR （超收敛分片恢复）超收敛解的自适应分析方法,与有限元最优网格进行了比较分析,进而提出反映自适应有限元收敛特性的估计式,并给出了自适应收敛率β的定义。该文给出的数值试验表明:采用m次单元,对于解答光滑的问题,SPR法与EEP法均可有效用于自适应求解,其位移可按最大模获得m+1的自适应收敛率;对于奇异因子为α（<1）的奇异问题,SPR法失效,而基于EEP法的自适应求解,其位移按最大模可获得m+α的自适应收敛率,远高于α的常规有限元收敛率。     

二维有限元线法自适应分析的若干新进展

有限元线法(FEMOL)是一种优良的半解析、半离散方法,将其比拟为广义一维问题,遂可将一维有限元中十分成功的单元能量投影(EEP)超收敛算法以及基于该法的自适应求解方法推广到二维有限元线法分析中,至今已在二维Poisson方程和弹性力学平面问题中取得了令人满意的进展.该文旨在报道这些进展和成果.该文简要介绍了线法的EE...     

一个高效的一维有限元自适应求解的新方案 第十三届全国结构工程学术大会特邀报告

基于新近提出的一维有限元后处理超收敛算法——单元能量投影(EEP)法,将有限元自适应求解问题转化为对超收敛解答的自适应分段多项式插值问题,一步便可获得最优的有限元网格划分,在该网格上再次进行有限元计算,即可获得满足用户给定的误差限的有限元解答。该法简单实用、快速高效,是一个颇具优势和潜力的自适应方法。文中以二阶常微分方程模型问题为例,对该法的形成思路和实施策略做一介绍,并给出有代表性的数值算例用以展示该法的优良性能和效果。     

一维变分不等式问题的自适应有限元分析新探

结构工程中的弹性薄膜接触和杆件弹塑性扭转等问题是典型的变分不等式问题,对其高效精确求解,特别是满足给定精度要求下的自适应求解,是挑战性课题。该文作者新近成功实现了一维变分不等式问题的自适应有限元分析,该文对此进展作一报道。对于变分不等式的有限元求解,该文提出区域二分法和C检验技术,极大提升了松弛迭代的收敛速度,一般4次~5次线性解即可得到收敛的有限元解答,进而采用作者提出的EEP(单元能量投影)超收敛公式计算超收敛解答,用其检验误差并指导网格细分,逐步得到堪称为数值精确解的解答,亦即得到按照最大模度量逐点满足精度要求的解答。该文给出的数值算例表明所提出的算法具有高效、可靠、精确的优良特性。     

二维自由振动问题的自适应有限元分析初探

自由振动反映结构动力特性,是抗震分析和结构设计的重要基础。近年来,基于单元能量投影（EEP）法的自适应有限元分析已在一系列线弹性及非线性问题中取得成功,而有限元线法（FEMOL）自适应分析在二维自由振动问题中的应用也被证实是有效的。在此基础上,该文进一步提出二维自由振动问题的自适应有限元分析方法。通过将特征值问题线性化,合理引入二维线性问题的EEP超收敛计算和自适应求解技术,该法可得到满足精度要求的自振频率和按最大模度量满足用户给定误差限的振型。该文以弹性薄膜为例,介绍了这一进展,并给出数值算例以表明该方法的有效性和可靠性。     

固端法:二维有限元先验定量误差估计与控制

该文基于有限元超收敛计算的单元能量投影(Element Energy Projection,简称EEP)法,尝试将一维有限元中新近提出的先验定量误差估计的“固端法”拓展到二维有限元分析,以Poisson方程为例,用EEP公式预先估算出各单元的误差,可以不经有限元求解计算而直接给出满足精度要求的网格划分。该文给出的初步数值算例验证了该法的有效性。     

运动方程时程单元先验步长估计初探

基于单元能量投影(element energy projection,EEP)法和边值问题固端法的思想,将其扩展至运动方程问题。该文以单自由度线性元为例,采用Taylor级数渐近展开,对问题的求解进行实质性简化计算;探讨了不经有限元求解便可进行先验定量误差估计的算法;进而实现了自适应单元步长的先验估计和确定。该文给出初步算例,验证了该方法的可行性和有效性。     

基于EEP技术的一维有限元结点位移误差计算

Using super-convergent solutions calculated by the Element Energy Projection (EEP) method, equivalent nodal load vectors from the residual load term were derived in this paper without changing the finite element (FE) meshes and the global stiffness matrices. The subsequent back-substitutions can generate highly accurate estimates for the errors of nodal displacements and hence greatly improve the nodal accuracy. Taking a general second-order ordinary differential equation as the model problem, the algorithm of the proposed method and associated numerical examples were given to show that the proposed method is simple and effective, and that using elements of degree m≥1, the improved nodal displacements can gain the super-super-convergence orders h2m+2 and h3m+mod(m, 2) for simplified and condensed EEP forms, respectively. A variety of significant further extensions and applications were also discussed.     

Numerical implementation of a neural network based material model in finite element analysis

Neural network (NN) based constitutive models can capture non‐linear material behaviour. These models are versatile and have the capacity to continuously learn as additional material response data becomes available. NN constitutive models are increasingly used within the finite element (FE) method for the solution of boundary value problems. NN constitutive models, unlike commonly used plasticity models, do not require special integration procedures for implementation in FE analysis. NN constitutive model formulation does not use a material stiffness matrix concept in contrast to the elasto‐plastic matrix central to conventional plasticity based models. This paper addresses numerical implementation issues related to the use of NN constitutive models in FE analysis. A consistent material stiffness matrix is derived for the NN constitutive model that leads to efficient convergence of the FE Newton iterations. The proposed stiffness matrix is general and valid regardless of the material behaviour represented by the NN constitutive model. Two examples demonstrate the performance of the proposed NN constitutive model implementation. Copyright © 2004 John Wiley & Sons, Ltd.     

Model updating of a rotating machine using the self-adaptive differential evolution algorithm

Despite the good accuracy of finite element (FE) models to represent the dynamic behaviour of mechanical systems, practical applications show significant discrepancies between analytical predictions and experimental results, which are mostly due to uncertainties on the geometry configuration, imprecise material parameters and vague boundary conditions. Thereby, different approaches have been proposed to solve the inverse problems associated with the updating of FE models. Among them, the techniques based on minimization processes have shown to be some of the most promising ones. In this paper, a self-adaptive heuristic optimization method, namely the self-adaptive differential evolution (SADE), is evaluated. Differently from the canonical differential evolution (DE) algorithm, the SADE strategy is able to update dynamically the required parameters such as population size, crossover parameter and perturbation rate. This is done by considering a defined convergence rate on the evolution process of the algorithm in order to reduce the number of evaluations of the objective function. For illustration purposes, the SADE strategy is applied to the solution of typical mathematical functions. Additionally, the strategy is equally used to update the FE model of a rotating machine composed by a horizontal flexible shaft, two rigid discs and two unsymmetrical bearings. For comparison purposes, the canonical DE is also used. The results indicate that the SADE algorithm is a recommended technique for dealing with this kind of inverse problem.