Show simple item record

Data-driven wall models for Reynolds Averaged Navier-Stokes simulations

hal.structure.identifierDAAA, ONERA, Université Paris Saclay [Meudon]
hal.structure.identifierModeling Enablers for Multi-PHysics and InteractionS [MEMPHIS]
dc.contributor.authorROMANELLI, Michele
hal.structure.identifierDAAA, ONERA, Université Paris Saclay [Meudon]
dc.contributor.authorBENEDDINE, Samir
hal.structure.identifierDAAA, ONERA, Université Paris-Saclay [Châtillon]
dc.contributor.authorMARY, Ivan
hal.structure.identifierCertified Adaptive discRete moDels for robust simulAtions of CoMplex flOws with Moving fronts [CARDAMOM]
hal.structure.identifierInstitut de Mathématiques de Bordeaux [IMB]
dc.contributor.authorBEAUGENDRE, Heloise
hal.structure.identifierInstitut de Mathématiques de Bordeaux [IMB]
hal.structure.identifierModeling Enablers for Multi-PHysics and InteractionS [MEMPHIS]
dc.contributor.authorBERGMANN, Michel
hal.structure.identifierDAAA, ONERA, Université Paris Saclay [Meudon]
dc.contributor.authorSIPP, Denis
dc.date.issued2023-01-02
dc.identifier.issn0142-727X
dc.description.abstractEnThis article presents a data-based methodology to build Reynolds-Averaged Navier-Stokes (RANS) wall models for aerodynamic simulations at low Mach numbers. Like classical approaches, the model is based on nondimensional local quantities derived from the wall friction velocity u τ , the wall viscosity µ w , and the wall density ρ w. A fully-connected neural network approximates the relation u + = f (y + , p +). We consider reference data (obtained with RANS simulations based on fine meshes up to the wall) of attached turbulent flows at various Reynolds numbers over different geometries of bumps, covering a range of wall pressure gradients. After training the neural networks on a subset of the reference data, the paper assesses their ability to accurately recover data for unseen conditions on meshes that have been trimmed from the wall up to an interface height where the learned wall law is applied. The network's interpolation and extrapolation capabilities are quantified and carefully examined. Overall, when tested within its interpolation and extrapolation capabilities, the neural network model shows good robustness and accuracy. The global error on the skin friction coefficient is a few percent and behaves consistently over all the considered test cases.
dc.language.isoen
dc.publisherElsevier
dc.subject.enwall model machine learning RANS neural network
dc.subject.enwall model
dc.subject.enmachine learning
dc.subject.enRANS
dc.subject.enneural network
dc.title.enData-driven wall models for Reynolds Averaged Navier-Stokes simulations
dc.typeArticle de revue
dc.identifier.doi10.1016/j.ijheatfluidflow.2022.109097
dc.subject.halPhysique [physics]/Mécanique [physics]/Mécanique des fluides [physics.class-ph]
bordeaux.journalInternational Journal of Heat and Fluid Flow
bordeaux.page109097
bordeaux.volume99
bordeaux.peerReviewedoui
hal.identifierhal-03918157
hal.version1
hal.popularnon
hal.audienceInternationale
hal.origin.linkhttps://hal.archives-ouvertes.fr//hal-03918157v1
bordeaux.COinSctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.jtitle=International%20Journal%20of%20Heat%20and%20Fluid%20Flow&rft.date=2023-01-02&rft.volume=99&rft.spage=109097&rft.epage=109097&rft.eissn=0142-727X&rft.issn=0142-727X&rft.au=ROMANELLI,%20Michele&BENEDDINE,%20Samir&MARY,%20Ivan&BEAUGENDRE,%20Heloise&BERGMANN,%20Michel&rft.genre=article


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record