Aarhus Universitets segl

English

Du er her: AU » Om AU » Organisation » Predicting drag on rough surfaces by transfer learning of...

Om AU

Predicting drag on rough surfaces by transfer learning of empirical correlations

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avis › Tidsskriftartikel › Forskning › peer review

Standard

Predicting drag on rough surfaces by transfer learning of empirical correlations. / Lee, Sangseung; Yang, Jiasheng ; Forooghi, Pourya et al.
I: Journal of Fluid Mechanics, Bind 933, A18, 25.02.2022.

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avis › Tidsskriftartikel › Forskning › peer review

Harvard

Lee, S, Yang, J, Forooghi, P, Stroh, A & Bagheri, S 2022, 'Predicting drag on rough surfaces by transfer learning of empirical correlations', Journal of Fluid Mechanics, bind 933, A18. https://doi.org/10.1017/jfm.2021.1041

APA

Lee, S., Yang, J., Forooghi, P., Stroh, A., & Bagheri, S. (2022). Predicting drag on rough surfaces by transfer learning of empirical correlations. Journal of Fluid Mechanics, 933, [A18]. https://doi.org/10.1017/jfm.2021.1041

CBE

Lee S, Yang J, Forooghi P, Stroh A, Bagheri S. 2022. Predicting drag on rough surfaces by transfer learning of empirical correlations. Journal of Fluid Mechanics. 933:Article A18. https://doi.org/10.1017/jfm.2021.1041

MLA

Lee, Sangseung et al. "Predicting drag on rough surfaces by transfer learning of empirical correlations". Journal of Fluid Mechanics. 2022. 933. https://doi.org/10.1017/jfm.2021.1041

Vancouver

Lee S, Yang J, Forooghi P, Stroh A, Bagheri S. Predicting drag on rough surfaces by transfer learning of empirical correlations. Journal of Fluid Mechanics. 2022 feb. 25;933:A18. doi: 10.1017/jfm.2021.1041

Author

Lee, Sangseung ; Yang, Jiasheng ; Forooghi, Pourya et al. / Predicting drag on rough surfaces by transfer learning of empirical correlations. I: Journal of Fluid Mechanics. 2022 ; Bind 933.

Bibtex

@article{bfa2d3cd85754e4093eec3de8d37f51d,
title = "Predicting drag on rough surfaces by transfer learning of empirical correlations",
abstract = "Recent developments in neural networks have shown the potential of estimating drag on irregular rough surfaces. Nevertheless, the difficulty of obtaining a large high-fidelity dataset to train neural networks is deterring their use in practical applications. In this study, we propose a transfer learning framework to model the drag on irregular rough surfaces even with a limited amount of direct numerical simulations. We show that transfer learning of empirical correlations, reported in the literature, can significantly improve the performance of neural networks for drag prediction. This is because empirical correlations include 'approximate knowledge' of the drag dependency in high-fidelity physics. The 'approximate knowledge' allows neural networks to learn the surface statistics known to affect drag more efficiently. The developed framework can be applied to applications where acquiring a large dataset is difficult but empirical correlations have been reported.",
keywords = "machine learning",
author = "Sangseung Lee and Jiasheng Yang and Pourya Forooghi and Alexander Stroh and Shervin Bagheri",
year = "2022",
month = feb,
day = "25",
doi = "10.1017/jfm.2021.1041",
language = "English",
volume = "933",
journal = "Journal of Fluid Mechanics",
issn = "0022-1120",
publisher = "Cambridge University Press",

}

RIS

TY - JOUR

T1 - Predicting drag on rough surfaces by transfer learning of empirical correlations

AU - Lee, Sangseung

AU - Yang, Jiasheng

AU - Forooghi, Pourya

AU - Stroh, Alexander

AU - Bagheri, Shervin

PY - 2022/2/25

Y1 - 2022/2/25

N2 - Recent developments in neural networks have shown the potential of estimating drag on irregular rough surfaces. Nevertheless, the difficulty of obtaining a large high-fidelity dataset to train neural networks is deterring their use in practical applications. In this study, we propose a transfer learning framework to model the drag on irregular rough surfaces even with a limited amount of direct numerical simulations. We show that transfer learning of empirical correlations, reported in the literature, can significantly improve the performance of neural networks for drag prediction. This is because empirical correlations include 'approximate knowledge' of the drag dependency in high-fidelity physics. The 'approximate knowledge' allows neural networks to learn the surface statistics known to affect drag more efficiently. The developed framework can be applied to applications where acquiring a large dataset is difficult but empirical correlations have been reported.

AB - Recent developments in neural networks have shown the potential of estimating drag on irregular rough surfaces. Nevertheless, the difficulty of obtaining a large high-fidelity dataset to train neural networks is deterring their use in practical applications. In this study, we propose a transfer learning framework to model the drag on irregular rough surfaces even with a limited amount of direct numerical simulations. We show that transfer learning of empirical correlations, reported in the literature, can significantly improve the performance of neural networks for drag prediction. This is because empirical correlations include 'approximate knowledge' of the drag dependency in high-fidelity physics. The 'approximate knowledge' allows neural networks to learn the surface statistics known to affect drag more efficiently. The developed framework can be applied to applications where acquiring a large dataset is difficult but empirical correlations have been reported.

KW - machine learning

U2 - 10.1017/jfm.2021.1041

DO - 10.1017/jfm.2021.1041

M3 - Journal article

VL - 933

JO - Journal of Fluid Mechanics

JF - Journal of Fluid Mechanics

SN - 0022-1120

M1 - A18

ER -