2004
DOI: 10.1016/s0927-5452(04)80033-9
|View full text |Cite
|
Sign up to set email alerts
|

Multilevel extended algorithms in structural dynamics on parallel computers

Abstract: A parallelization concept for the adaptive multi-level substructuring (AMLS) method is presented, the idea of which is to hierarchically substructure the system under consideration, and at the same time to use a truncated eigenvalue decomposition on the interfaces to reduce the excessive number of interface degrees of freedom in component mode methods.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1

Citation Types

0
4
0

Year Published

2006
2006
2024
2024

Publication Types

Select...
5

Relationship

0
5

Authors

Journals

citations
Cited by 5 publications
(4 citation statements)
references
References 7 publications
0
4
0
Order By: Relevance
“…5 The relative error of eigenvalues for the different number of components compared to the eigenvalues obtained from the block Lanczos method in the plate type FE model (ss ¼ 64, 128, 256 components) is opposite of step 1 because the size of the transformed matrix increases as the number of components grows. Generally, with the same component cutoff frequency, the number of component modes, s, for forming the transformation matrix increases as the number of components increases or the size of components decreases [8,9], which results in the dimension of the transformed matrices increases. However, the accuracy of eigenvalues decreases as the number of components increases with the same component cutoff frequency a ¼ 5, shown in Table 3.…”
Section: Example 1: Shell Type Fe Modelmentioning
confidence: 99%
See 1 more Smart Citation
“…5 The relative error of eigenvalues for the different number of components compared to the eigenvalues obtained from the block Lanczos method in the plate type FE model (ss ¼ 64, 128, 256 components) is opposite of step 1 because the size of the transformed matrix increases as the number of components grows. Generally, with the same component cutoff frequency, the number of component modes, s, for forming the transformation matrix increases as the number of components increases or the size of components decreases [8,9], which results in the dimension of the transformed matrices increases. However, the accuracy of eigenvalues decreases as the number of components increases with the same component cutoff frequency a ¼ 5, shown in Table 3.…”
Section: Example 1: Shell Type Fe Modelmentioning
confidence: 99%
“…Recently, multi-level substructuring method has been introduced [8,9]. In the multi-level substructuring method, an FE model is divided into a number of smaller components on multi-level contrary to traditional CMS methods, which are single level.…”
Section: Introductionmentioning
confidence: 99%
“…The implementation of CMS method or substructuring method on multi-level has been attempted [8][9][10][11][12][13][14][15] as a generalization of the classical CMS method. Since most of researches used in-core approach, which uses a computer's main memory in numerical implementation, it is strongly required to use out-of-core approach in order to solve practical large scale industry applications.…”
Section: Introductionmentioning
confidence: 99%
“…Recently, the recursive component mode synthesis (RCMS) method or multi-level substructuring method has been introduced [7][8][9]. The RCMS method is quite effective for large-scale structural systems [9].…”
Section: Introductionmentioning
confidence: 99%