Data-driven augmentation of a RANS turbulence model for transonic flow prediction

Cornelia Grabe; Florian Jäckel; Parv Khurana; Richard P. Dwight

doi:10.1108/HFF-08-2022-0488

Data-driven augmentation of a RANS turbulence model for transonic flow prediction

Cornelia Grabe^*, Florian Jäckel, Parv Khurana, Richard P. Dwight

^*Corresponding author for this work

Aerodynamics

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

1 Citation (Scopus)

43 Downloads (Pure)

Abstract

Purpose: This paper aims to improve Reynolds-averaged Navier Stokes (RANS) turbulence models using a data-driven approach based on machine learning (ML). A special focus is put on determining the optimal input features used for the ML model. Design/methodology/approach: The field inversion and machine learning (FIML) approach is applied to the negative Spalart-Allmaras turbulence model for transonic flows over an airfoil where shock-induced separation occurs. Findings: Optimal input features and an ML model are developed, which improve the existing negative Spalart-Allmaras turbulence model with respect to shock-induced flow separation. Originality/value: A comprehensive workflow is demonstrated that yields insights on which input features and which ML model should be used in the context of the FIML approach.

Original language	English
Pages (from-to)	1544-1561
Number of pages	18
Journal	International Journal of Numerical Methods for Heat and Fluid Flow
Volume	33
Issue number	4
DOIs	https://doi.org/10.1108/HFF-08-2022-0488
Publication status	Published - 2023

Bibliographical note

Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care
Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Keywords

Data-driven turbulence modeling
Feature selection
Flow separation
Machine learning
RANS
Transonic flows

Access to Document

10.1108/HFF-08-2022-0488

10-1108_HFF-08-2022-0488Final published version, 974 KB

Cite this

@article{8db07b554656443db08f695affda1cd0,

title = "Data-driven augmentation of a RANS turbulence model for transonic flow prediction",

abstract = "Purpose: This paper aims to improve Reynolds-averaged Navier Stokes (RANS) turbulence models using a data-driven approach based on machine learning (ML). A special focus is put on determining the optimal input features used for the ML model. Design/methodology/approach: The field inversion and machine learning (FIML) approach is applied to the negative Spalart-Allmaras turbulence model for transonic flows over an airfoil where shock-induced separation occurs. Findings: Optimal input features and an ML model are developed, which improve the existing negative Spalart-Allmaras turbulence model with respect to shock-induced flow separation. Originality/value: A comprehensive workflow is demonstrated that yields insights on which input features and which ML model should be used in the context of the FIML approach.",

keywords = "Data-driven turbulence modeling, Feature selection, Flow separation, Machine learning, RANS, Transonic flows",

author = "Cornelia Grabe and Florian J{\"a}ckel and Parv Khurana and Dwight, {Richard P.}",

note = "Green Open Access added to TU Delft Institutional Repository {\textquoteleft}You share, we take care!{\textquoteright} – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public. ",

year = "2023",

doi = "10.1108/HFF-08-2022-0488",

language = "English",

volume = "33",

pages = "1544--1561",

journal = "International Journal of Numerical Methods for Heat and Fluid Flow",

issn = "0961-5539",

publisher = "Emerald Publishing",

number = "4",

}

TY - JOUR

T1 - Data-driven augmentation of a RANS turbulence model for transonic flow prediction

AU - Grabe, Cornelia

AU - Jäckel, Florian

AU - Khurana, Parv

AU - Dwight, Richard P.

N1 - Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2023

Y1 - 2023

N2 - Purpose: This paper aims to improve Reynolds-averaged Navier Stokes (RANS) turbulence models using a data-driven approach based on machine learning (ML). A special focus is put on determining the optimal input features used for the ML model. Design/methodology/approach: The field inversion and machine learning (FIML) approach is applied to the negative Spalart-Allmaras turbulence model for transonic flows over an airfoil where shock-induced separation occurs. Findings: Optimal input features and an ML model are developed, which improve the existing negative Spalart-Allmaras turbulence model with respect to shock-induced flow separation. Originality/value: A comprehensive workflow is demonstrated that yields insights on which input features and which ML model should be used in the context of the FIML approach.

AB - Purpose: This paper aims to improve Reynolds-averaged Navier Stokes (RANS) turbulence models using a data-driven approach based on machine learning (ML). A special focus is put on determining the optimal input features used for the ML model. Design/methodology/approach: The field inversion and machine learning (FIML) approach is applied to the negative Spalart-Allmaras turbulence model for transonic flows over an airfoil where shock-induced separation occurs. Findings: Optimal input features and an ML model are developed, which improve the existing negative Spalart-Allmaras turbulence model with respect to shock-induced flow separation. Originality/value: A comprehensive workflow is demonstrated that yields insights on which input features and which ML model should be used in the context of the FIML approach.

KW - Data-driven turbulence modeling

KW - Feature selection

KW - Flow separation

KW - Machine learning

KW - RANS

KW - Transonic flows

UR - http://www.scopus.com/inward/record.url?scp=85152915315&partnerID=8YFLogxK

U2 - 10.1108/HFF-08-2022-0488

DO - 10.1108/HFF-08-2022-0488

M3 - Article

AN - SCOPUS:85152915315

SN - 0961-5539

VL - 33

SP - 1544

EP - 1561

JO - International Journal of Numerical Methods for Heat and Fluid Flow

JF - International Journal of Numerical Methods for Heat and Fluid Flow

IS - 4

ER -

Data-driven augmentation of a RANS turbulence model for transonic flow prediction

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this