A feature points-based method for data transfer in fluid-structure interactions. (15th November 2022)
- Record Type:
- Journal Article
- Title:
- A feature points-based method for data transfer in fluid-structure interactions. (15th November 2022)
- Main Title:
- A feature points-based method for data transfer in fluid-structure interactions
- Authors:
- DOU, Weiyuan
GUO, Sheng
ZHANG, Lele
ZHU, Yu
STICHEL, Sebastian - Abstract:
- Highlights: A feature points-based combined interpolation method for non-matching meshes data transfer in FSI is proposed. The combined interpolation is realized by means of Simpson's rule. The influence of the weigh coefficient is investigated by comparative studies. Applicability and improvement on data transfer accuracy are validated using a series of examples. Abstract: For numerical simulations of fluid-structure interaction (FSI) problems, discretized meshes of different computational domains do not have to match at the common interface. Data transfer via the interface is therefore significantly affecting the accuracy of FSI. Implementing data transfer between such pairwise non-matching meshes is challenging, as it is highly desirable to yield numerical accuracy using methods such as interpolation or projection. To further improve the data transfer accuracy for mesh-based approaches, this study proposes a feature points-based combined interpolation method by means of Simpson's rule, with which an appropriate weighting coefficient is determined. Moreover, radial basis function (RBF) is employed to perform the interpolation of scattered data points. The proposed method is validated by a series of examples involving various non-matching mesh configurations in FSI problems. The numerical findings indicate that this method can further improve data transfer accuracy. Even for a critical case with large mesh ratio at the interface it can provide satisfying results. GraphicalHighlights: A feature points-based combined interpolation method for non-matching meshes data transfer in FSI is proposed. The combined interpolation is realized by means of Simpson's rule. The influence of the weigh coefficient is investigated by comparative studies. Applicability and improvement on data transfer accuracy are validated using a series of examples. Abstract: For numerical simulations of fluid-structure interaction (FSI) problems, discretized meshes of different computational domains do not have to match at the common interface. Data transfer via the interface is therefore significantly affecting the accuracy of FSI. Implementing data transfer between such pairwise non-matching meshes is challenging, as it is highly desirable to yield numerical accuracy using methods such as interpolation or projection. To further improve the data transfer accuracy for mesh-based approaches, this study proposes a feature points-based combined interpolation method by means of Simpson's rule, with which an appropriate weighting coefficient is determined. Moreover, radial basis function (RBF) is employed to perform the interpolation of scattered data points. The proposed method is validated by a series of examples involving various non-matching mesh configurations in FSI problems. The numerical findings indicate that this method can further improve data transfer accuracy. Even for a critical case with large mesh ratio at the interface it can provide satisfying results. Graphical Abstract: Image, graphical abstract … (more)
- Is Part Of:
- International journal of mechanical sciences. Volume 234(2022)
- Journal:
- International journal of mechanical sciences
- Issue:
- Volume 234(2022)
- Issue Display:
- Volume 234, Issue 2022 (2022)
- Year:
- 2022
- Volume:
- 234
- Issue:
- 2022
- Issue Sort Value:
- 2022-0234-2022-0000
- Page Start:
- Page End:
- Publication Date:
- 2022-11-15
- Subjects:
- Fluid-structure interaction (FSI) -- Simpson's rule -- Feature points-based combined interpolation -- Data transfer -- Non-matching meshes
Mechanical engineering -- Periodicals
Génie mécanique -- Périodiques
Mechanical engineering
Maschinenbau
Mechanik
Zeitschrift
Periodicals
621.05 - Journal URLs:
- http://www.sciencedirect.com/science/journal/00207403 ↗
http://www.elsevier.com/journals ↗ - DOI:
- 10.1016/j.ijmecsci.2022.107696 ↗
- Languages:
- English
- ISSNs:
- 0020-7403
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 4542.344000
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 24114.xml