期刊
COMPUTERS & STRUCTURES
卷 84, 期 19-20, 页码 1190-1199出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.compstruc.2006.01.040
关键词
wavelet analysis; multiresolutional techniques; composite materials; finite element method; vibration analysis; transient heat transfer
The main goal of the paper is to use all the opportunities created by the wavelet analysis to model interface defects and some structural non-homogeneities in unidirectional periodic composite materials. The influence of these defects is detected by the comparison of a defected composite with the behaviour of additional composites with perfect interfaces. Some algebraic combination of the Haar wavelet, Mexican hat and harmonic wavelets is proposed to describe analytically the material properties, i.e. heat conductivity and capacity, Young modulus as well as mass density. The interface defects are simulated here as a significant decrease of any material property in the neighbourhood of a multi-material boundary inserted into the component having smaller material coefficients. The composites with and without interface defects with material characteristics so defined are compared with one another using some basic wavelet-based FEM tests of different discretisation order for transient heat transfer and free vibrations problems. The multiresolutional meshing process is completed thanks to various order meshes, where 1D two- or three-noded finite elements discretize a single representative volume element (RVE). Necessary numerical data pre-processing is carried out using the symbolic computations package MAPLE, which was developed for determination of the effective characteristics for analogous composites. Computational studies highlight the crucial role of the interface in the macro-scale behaviour of even unidirectional composites and should be extended to multiresolutional elastoplastic computational analysis, 2D and 3D analysis by the wavelet functions. (c) 2006 Civil-Comp Ltd. and Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据