Unpublished conference/Abstract (Scientific congresses, symposiums and conference proceedings)
On the Smoothed eXtended Finite Element Method for Continuum
Natarajan, Sundararajan; BORDAS, Stéphane; Rabczuk, Timon et al.
200917th. UK Conference on Computational Mechanics (ACME-UK)
 

Files


Full Text
20090318_SmXFEM_ACME09.pdf
Author preprint (135.27 kB)
Request a copy
Annexes
20090402ACME09_SmXFEM.pdf
(474.44 kB)
Request a copy

All documents in ORBilu are protected by a user license.

Send to



Details



Keywords :
Strain smoothing; XFEM; Smoothed extended finite element method; numerical integration
Abstract :
[en] In this paper, we combine the strain smoothing technique proposed by Liu et al [1] coined as the smoothed finite element method (SFEM) to partition of unity methods, namely the extended finite element method (XFEM) [2] to give birth to the smoothed extended finite element method (SmXFEM) [3]. SmXFEM shares properties both with the SFEM and the XFEM. The proposed method eliminates the need to compute and integrate the derivatives of shape functions (which are singular at the tip for linear elastic fracture mechanics). The need for isoparametric mapping is eliminated because the integration is done along the boundary of the finite element or smoothing cells, which allows elements of arbitrary shape. We present numerical results for various differential equations that have singularity or steep gradient at the boundary. The method is verified on several examples and comparisons are made to the conventional XFEM.
Disciplines :
Aerospace & aeronautics engineering
Author, co-author :
Natarajan, Sundararajan
BORDAS, Stéphane ;  University of Luxembourg > Faculty of Science, Technology and Communication (FSTC) > Engineering Research Unit
Rabczuk, Timon
Guo, Zaoyang
Language :
English
Title :
On the Smoothed eXtended Finite Element Method for Continuum
Publication date :
April 2009
Event name :
17th. UK Conference on Computational Mechanics (ACME-UK)
Event date :
6-8 April 2009
Audience :
International
Focus Area :
Computational Sciences
Available on ORBilu :
since 27 November 2013

Statistics


Number of views
110 (2 by Unilu)
Number of downloads
0 (0 by Unilu)

Bibliography


Similar publications



Contact ORBilu