2018
DOI: 10.1016/j.compstruc.2018.04.007
|View full text |Cite
|
Sign up to set email alerts
|

Solving generalized eigenvalue problems for large scale fluid-structure computational models with mid-power computers

Abstract: This article proposes a method for solving generalized eigenvalue problems on medium-power computers with a moderate memory in the particular context of studying fluid-structure systems with sloshing and capillarity. This research was performed following many RAM problems encountered when computing the modal characterization of the system studied. The methodology proposed is one solution to reduce RAM and time required for the computation, by using methods such as double projection or subspace iterations.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
2

Citation Types

0
5
0

Year Published

2019
2019
2022
2022

Publication Types

Select...
4
1

Relationship

3
2

Authors

Journals

citations
Cited by 5 publications
(5 citation statements)
references
References 32 publications
0
5
0
Order By: Relevance
“…For large-scale 3D computational models, solving the three generalized eigenvalue problems with standard algorithms induced major difficulties for mid-power computers. This is the reason why non-standard algorithms have been proposed in (Akkaoui et al, 2018) for solving these generalized eigenvalue problems. The limitations of the computer resources are principally due to the RAM limitations and a prohibitive CPU time.…”
Section: Construction Of the Nonlinear Reduced-order Computational Modelmentioning
confidence: 99%
See 1 more Smart Citation
“…For large-scale 3D computational models, solving the three generalized eigenvalue problems with standard algorithms induced major difficulties for mid-power computers. This is the reason why non-standard algorithms have been proposed in (Akkaoui et al, 2018) for solving these generalized eigenvalue problems. The limitations of the computer resources are principally due to the RAM limitations and a prohibitive CPU time.…”
Section: Construction Of the Nonlinear Reduced-order Computational Modelmentioning
confidence: 99%
“…In addition, the computation of the reducedorder bases exhibits major difficulties in terms of computational cost when studying a large-scale fluid-structure model due (i) to the high flexural motion of the thin-walled cylinder and (ii) to the free surface. These difficulties are circumvented by using the algorithms presented in (Akkaoui et al, 2018). The outline of the paper is as follows.…”
Section: Introductionmentioning
confidence: 99%
“…For solving the three generalized eigenvalue problems for large scale fluid-structure computational models with mid-power computers, we refer the reader to [63] in which all the details are given, for which the matrices of the generalized eigenvalue problems must be those introduced in Section 5.3.1, that is to say, using quadratic finite elements.…”
Section: Reduced-order Computational Model For the Coupled Systemmentioning
confidence: 99%
“…Such computation on mid-power computers can be very challenging when large finite element meshes are involved. The original computational strategy [5] that allows for circumventing these difficulties is used in this analysis. The nonlinear reduced-order model is then written as…”
Section: Description Of the Computational Modelmentioning
confidence: 99%