首页> 外文会议>Computational Mechanics >A Regularized Algorithm for the Calculation of Second Derivative Values in Potential BEM
【24h】

A Regularized Algorithm for the Calculation of Second Derivative Values in Potential BEM

机译:潜在BEM中二阶导数计算的正则算法

获取原文

摘要

In the past decade, experiments have shown that materials display strong size effects when the characteristic length scale is on the order of microns. The conventional elasto-plasticity theory, however, can not predict this size dependence because its constitutive model possesses no internal length scale. So far, a number of models considering this length scale have been proposed, see e.g. [1], and numerical simulations have been made to this phenomenon through the finite element method (FEM) [2] and the boundary element method (BEM) [3]. In those numerical analyses, strain gradients were generally calculated in an indirect way, i.e. interpolating from the obtained strains or interpolating twice from the displacements. This procedure obviously induces significant numerical errors. Therefore, the aim of the present work is to develop a new algorithm to improve the numerical accuracy of the strain gradients. For the first step, this paper presents our new algorithm in the context of potential BEM to show how the potential second derivatives can be calculated in an agreeable accuracy.The boundary integral equations (BIEs) of potential second derivatives are of third order singularities; here the strong singularity is referred to as the first order singularity and the hyper-singularity as the second order one. Obviously the direct calculation of the third order singular integrals is rather difficult, which has been seldom met in the BEM analysis so far. The idea of the present paper is to use a special indirect algorithm which is based on a regularized BIE formulation of the potential second derivatives, following the work of the present author and his coworkers [4]. In this algorithm, the singularities in the integrals are fully regularized, thus only a Gaussian quadrature is needed and the most common used C~0 continuous element can be used. Numerical examples show the accuracy and efficiency of the present algorithm.
机译:在过去的十年中,实验表明,当特征长度标度为微米量级时,材料会显示出强大的尺寸效应。然而,传统的弹塑性理论不能预测这种尺寸依赖性,因为其本构模型不具有内部长度尺度。到目前为止,已经提出了许多考虑该长度尺度的模型,例如参见。 [1],并通过有限元方法(FEM)[2]和边界元方法(BEM)[3]对这种现象进行了数值模拟。在那些数值分析中,通常以间接的方式来计算应变梯度,即从获得的应变进行内插或从位移进行两次内插。该过程显然会引起明显的数值误差。因此,本工作的目的是开发一种新的算法来提高应变梯度的数值精度。第一步,本文在潜在BEM的背景下展示了我们的新算法,以展示如何以合理的精度计算潜在的二阶导数。 潜在的二阶导数的边界积分方程(BIE)具有三阶奇异性。在这里,强奇异性称为一阶奇异性,超奇异性称为二阶奇异性。显然,直接计算三阶奇异积分是相当困难的,到目前为止,在BEM分析中很少遇到。本文的想法是使用一种特殊的间接算法,该算法基于作者和他的同事[4]的工作,基于潜在二阶导数的正则化BIE公式[4]。在该算法中,积分中的奇点被完全正则化,因此只需要一个高斯正交,就可以使用最常用的C〜0连续元素。数值算例表明了该算法的准确性和有效性。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号