2016
DOI: 10.1007/s00466-016-1268-0
|View full text |Cite
|
Sign up to set email alerts
|

Convergence acceleration for partitioned simulations of the fluid-structure interaction in arteries

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
29
0

Year Published

2017
2017
2024
2024

Publication Types

Select...
7

Relationship

4
3

Authors

Journals

citations
Cited by 23 publications
(29 citation statements)
references
References 57 publications
0
29
0
Order By: Relevance
“…1, the parallel coupling algorithm in combination with the original QNLS method yields the lowest number of iterations. The variant FQNLSMTR, in which data from previous time steps is reused and where the filtering according to [4] and resetting according to [5] are applied, unexpectedly does not yield a performance gain. While this behavior was not observed in preliminary investigations [8], the fact that the Aitken relaxation-like method does not yield stable simulations in combination with the parallel coupling algorithms (marked as failed in Fig.…”
Section: Partitioned Couplingmentioning
confidence: 98%
See 1 more Smart Citation
“…1, the parallel coupling algorithm in combination with the original QNLS method yields the lowest number of iterations. The variant FQNLSMTR, in which data from previous time steps is reused and where the filtering according to [4] and resetting according to [5] are applied, unexpectedly does not yield a performance gain. While this behavior was not observed in preliminary investigations [8], the fact that the Aitken relaxation-like method does not yield stable simulations in combination with the parallel coupling algorithms (marked as failed in Fig.…”
Section: Partitioned Couplingmentioning
confidence: 98%
“…A parallel coupling algorithm can be based on the FPI Regarding the convergence acceleration method, various choices are possible. We have found that the quasi-Newton least squares (QNLS) method from [3] with the enhancements introduced in [4] and [5] yields the lowest number of coupling iterations for both coupling algorithms. However, less complex methods such as the Aitken relaxation-like approach used for fixed-point FSI simulations in [6] are reasonable alternatives in view of a lower implementation effort.…”
Section: Partitioned Couplingmentioning
confidence: 99%
“…Aitken's underrelaxation constants are adjusted dynamically based on the acceleration of the rigid body, thus improving stability. This type of simulations with deformable grids, applied to fluidstructure interaction using OpenFOAM solvers have been tested in the past with complex problems, such as drag reduction [38], flow and deformation in arteries [39], bluff body oscillations [40], or offshore floating platforms [41]).…”
Section: Numerical Simulationsmentioning
confidence: 99%
“…j+1,k using a suitable convergence acceleration scheme. In our experience, the quasi-Newton least-squares method proposed in Degroote et al (2009) or Radtke et al (2016) has proven effective for different kinds of strongly coupled multi-field problems. In this scheme, the displacement increment u h n,1 j+1,k is determined by minimising the updated residual…”
Section: Partitioned Solution Approachmentioning
confidence: 99%