1986
DOI: 10.1002/nme.1620230610
|View full text |Cite
|
Sign up to set email alerts
|

Analysis of spurious eigenmodes in finite element equations

Abstract: Numerical analysis of difference schemes often reveals the presence of eigenmodes which do not feature in the continuum solution. An examination of the dispersion relation shows how the spurious and physical modes interact. The behaviour of certain wave-profiles was predicted using this analysis and the results confirmed by numerical experiment. 'New address. British Gas Corporation, Engineering Research Station, Harvey Combe, Killingworth P.O. Box 1 LH, Newcastle-upon-Tyne, NE99 ILH.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2

Citation Types

0
2
0

Year Published

1990
1990
1995
1995

Publication Types

Select...
5

Relationship

0
5

Authors

Journals

citations
Cited by 6 publications
(2 citation statements)
references
References 5 publications
0
2
0
Order By: Relevance
“…26 One way of measuring the accuracy of a numerical scheme is to compare the modulus and argument of a numerical eigenvalue to those of the continuum eigenvalue. For the amplitude error define iku A t t , = e -.…”
Section: Fourier Analysismentioning
confidence: 99%
See 1 more Smart Citation
“…26 One way of measuring the accuracy of a numerical scheme is to compare the modulus and argument of a numerical eigenvalue to those of the continuum eigenvalue. For the amplitude error define iku A t t , = e -.…”
Section: Fourier Analysismentioning
confidence: 99%
“…Note that because one continuum equation is approximated by two discrete equations, one for the primary unknown and the other for its derivative, there are two eigenvalues, physical and computational, associated with a numerical scheme when the cubic element is used. 26 One way of measuring the accuracy of a numerical scheme is to compare the modulus and argument of a numerical eigenvalue to those of the continuum eigenvalue. For the amplitude error define Polar plots of the amplitude and phase errors are given in Figure 1, and the discrete L2-error (<), measured by (22) with a normalization factor of unity, is given in Table I: where y j is a discrete value of either R or 0, and N is the total number of discrete points.…”
Section: Fourier Analysismentioning
confidence: 99%