Standard

APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING. / Bernard, A; Yakovenko, S.

In: Eurasian Journal of Mathematical and Computer Applications, Vol. 13, No. 2, 30.06.2025, p. 4-12.

Research output: Contribution to journal › Article › peer-review

Harvard

Bernard, A & Yakovenko, S 2025, 'APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING', Eurasian Journal of Mathematical and Computer Applications, vol. 13, no. 2, pp. 4-12. https://doi.org/10.32523/2306-6172-2025-13-2-4-12

APA

Bernard, A., & Yakovenko, S. (2025). APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING. Eurasian Journal of Mathematical and Computer Applications, 13(2), 4-12. https://doi.org/10.32523/2306-6172-2025-13-2-4-12

Vancouver

Bernard A, Yakovenko S. APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING. Eurasian Journal of Mathematical and Computer Applications. 2025 Jun 30;13(2):4-12. doi: 10.32523/2306-6172-2025-13-2-4-12

Author

Bernard, A ; Yakovenko, S. / APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING. In: Eurasian Journal of Mathematical and Computer Applications. 2025 ; Vol. 13, No. 2. pp. 4-12.

BibTeX

@article{413d5891d0a840e084fcbe5be2088631,
title = "APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING",
abstract = "Machine-learning methods to enhance an approximation for the Reynolds-stress anisotropy tensor are presented. The approach of tensor basis random forest is applied for this. The set of input features and tensors in the basis are discussed. Different ways to propagate the Reynolds stress anisotropy tensor into the Reynolds-averaged Navier–Stokes equation solver are explored. It is demonstrated that the conventional expression for Reynolds-stress anisotropy based on the linear eddy-viscosity model is not able to reproduce a secondary flow in the square duct cross-section, whereas the machine-learning modifications can fix such a disadvantage and have potentials for further improvements of turbulence models.",
keywords = "tensor basis random forest, Reynolds stress, RANS, LES, DNS, channel flow",
author = "A Bernard and S Yakovenko",
note = " The study has been supported by a grant No. 22-19-00587 of Russian Science Foundation, https://rscf.ru/en/project/22-19-00587/ Bernard A., Yakovenko S. N. APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING / A. Bernard, S. N. Yakovenko // Eurasian Journal of Mathematical and Computer Applications. - 2025. - Т. 13. № 2. С. 4 - 12. DOI: 10.32523/2306-6172-2025-13-2-4-12 ",
year = "2025",
month = jun,
day = "30",
doi = "10.32523/2306-6172-2025-13-2-4-12",
language = "English",
volume = "13",
pages = "4--12",
journal = "Eurasian Journal of Mathematical and Computer Applications",
issn = "2306-6172",
publisher = "L. N. Gumilyov Eurasian National University",
number = "2",

}

RIS

TY - JOUR

T1 - APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING

AU - Bernard, A

AU - Yakovenko, S

N1 - The study has been supported by a grant No. 22-19-00587 of Russian Science Foundation, https://rscf.ru/en/project/22-19-00587/ Bernard A., Yakovenko S. N. APPLICATION OF MACHINE LEARNING METHODS TO CHANNEL FLOW MODELLING / A. Bernard, S. N. Yakovenko // Eurasian Journal of Mathematical and Computer Applications. - 2025. - Т. 13. № 2. С. 4 - 12. DOI: 10.32523/2306-6172-2025-13-2-4-12

PY - 2025/6/30

Y1 - 2025/6/30

N2 - Machine-learning methods to enhance an approximation for the Reynolds-stress anisotropy tensor are presented. The approach of tensor basis random forest is applied for this. The set of input features and tensors in the basis are discussed. Different ways to propagate the Reynolds stress anisotropy tensor into the Reynolds-averaged Navier–Stokes equation solver are explored. It is demonstrated that the conventional expression for Reynolds-stress anisotropy based on the linear eddy-viscosity model is not able to reproduce a secondary flow in the square duct cross-section, whereas the machine-learning modifications can fix such a disadvantage and have potentials for further improvements of turbulence models.

AB - Machine-learning methods to enhance an approximation for the Reynolds-stress anisotropy tensor are presented. The approach of tensor basis random forest is applied for this. The set of input features and tensors in the basis are discussed. Different ways to propagate the Reynolds stress anisotropy tensor into the Reynolds-averaged Navier–Stokes equation solver are explored. It is demonstrated that the conventional expression for Reynolds-stress anisotropy based on the linear eddy-viscosity model is not able to reproduce a secondary flow in the square duct cross-section, whereas the machine-learning modifications can fix such a disadvantage and have potentials for further improvements of turbulence models.

KW - tensor basis random forest

KW - Reynolds stress

KW - RANS

KW - LES

KW - DNS

KW - channel flow

UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=105016710333&origin=inward

U2 - 10.32523/2306-6172-2025-13-2-4-12

DO - 10.32523/2306-6172-2025-13-2-4-12

M3 - Article

VL - 13

SP - 4

EP - 12

JO - Eurasian Journal of Mathematical and Computer Applications

JF - Eurasian Journal of Mathematical and Computer Applications

SN - 2306-6172

IS - 2

ER -

ID: 69975078