Data-driven calibration of RANS heat transfer prediction on a curved rough surface

Ignatowicz, Kevin, Morency, François, Beaugendre, Héloïse et Solaï, Elie. 2022. « Data-driven calibration of RANS heat transfer prediction on a curved rough surface ». In 56th 3AF International Conference on Applied Aerodynamics (Toulouse, France, March 28-30, 2022)

Prévisualisation

PDF Morency-F-2022-24336.pdf - Version acceptée Licence d'utilisation : Tous les droits réservés aux détenteurs du droit d'auteur. Télécharger (863kB) | Prévisualisation

Résumé

For heat transfer predictions using RANS simulations, turbulence models require adjustments for rough surfaces. The drawback of these adjustments is the tendency to over predict the heat transfers compared to experiments. These over predictions require the use of an additional thermal correction model to lower the heat transfers. Inputting in the correction model the numerical parameters giving the best fit with experimental results is non-trivial, since actual roughness patterns are often irregular. The objective of the paper is to develop a methodology to calibrate two thermal correction models for a rough curved channel test case. First, a design of experiments of heat transfers is built, then metamodels are generated. Finally, the metamodels are used by a Bayesian inversion procedure estimating the best set of input parameters allowing fitting the experimental results. This methodology allows obtaining less than 7% of average discrepancy between the RANS prediction and the experimental results.

Type de document:	Compte rendu de conférence
Informations complémentaires:	Identifiant de l'article: FP70-AERO2022-ignatowicz
Professeur:	Professeur Morency, François
Affiliation:	Génie mécanique
Date de dépôt:	09 mai 2022 15:48
Dernière modification:	25 mai 2022 14:07
URI:	https://espace2.etsmtl.ca/id/eprint/24336

Actions (Authentification requise)

Dernière vérification avant le dépôt