Three-Dimensional Swirl Flow Velocity-Field Reconstruction Using a Neural Network With Radial Basis Functions - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Journal of Fluids Engineering Année : 2001

Three-Dimensional Swirl Flow Velocity-Field Reconstruction Using a Neural Network With Radial Basis Functions

Résumé

For many studies, knowledge of continuous evolution of hydrodynamic characteristics is useful but generally measurement techniques provide only discrete information. In the case of complex flows, usual numerical interpolating methods appear to be not adapted, as for the free decaying swirling flow presented in this study. The three-dimensional motion involved induces a spatial dependent velocity-field. Thus, the interpolating method has to be three-dimensional and to take into account possible flow nonlinearity, making common methods unsuitable. A different interpolation method is thus proposed, based on a neural network algorithm with Radial Basis Functions.
Fichier principal
Vignette du fichier
2001_Pruvost_Neural_JFluidEng.pdf (97.64 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02533987 , version 1 (07-04-2020)

Identifiants

Citer

Jeremy Pruvost, J. Legrand, P. Legentilhomme. Three-Dimensional Swirl Flow Velocity-Field Reconstruction Using a Neural Network With Radial Basis Functions. Journal of Fluids Engineering, 2001, 123 (4), pp.920-927. ⟨10.1115/1.1412847⟩. ⟨hal-02533987⟩
27 Consultations
84 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More