Please use this identifier to cite or link to this item:
Title: A radial boundary node method for two-dimensional elastic analysis
Authors: Xie, H.
Nogami, T. 
Wang, J. 
Keywords: Boundary integral equation
Boundary node method
Point interpolation method
Radial basis function
Shape parameter
Issue Date: Sep-2003
Citation: Xie, H., Nogami, T., Wang, J. (2003-09). A radial boundary node method for two-dimensional elastic analysis. Engineering Analysis with Boundary Elements 27 (8) : 853-862. ScholarBank@NUS Repository.
Abstract: The boundary node method (BNM) takes the advantages of both the boundary integral equation in dimension reduction and the moving least-square (MLS) approximation in elements elimination. However, the BNM inherits the deficiency of the MLS approximation, in which the shape functions lack the delta function property. As a result boundary conditions could not be exactly implemented for the BNM. In this paper, a radial boundary node method (RBNM) is proposed. The RBNM uses radial basis functions (RBFs) instead of the MLS to construct its interpolation. Consequently, the interpolation function could pass through nodes exactly, and the shape functions are of the delta function property. The exponential (EXP) and the multiquadric (MQ) RBFs are used in the current RBNM, and their shape parameters are studied in detail through some analyses of two-dimensional elastic problems. The suitable ranges of the shape parameters are proposed for both the EXP and the MQ basis functions. It is found that the RBNM is as accurate as or even more precise than the BEM. This suggests that the current RBNM could be robust and applicable. © 2003 Published by Elsevier Ltd.
Source Title: Engineering Analysis with Boundary Elements
ISSN: 09557997
DOI: 10.1016/S0955-7997(03)00027-4
Appears in Collections:Staff Publications

Show full item record
Files in This Item:
There are no files associated with this item.

Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.