ENGLISH
La vitrine de diffusion des publications et contributions des chercheurs de l'ÉTS
RECHERCHER

AI-driven prediction of central plug morphology in Poiseuille flow of Bingham fluids with superhydrophobic walls

Joulaei, Amir, Rahmani, Hossein et Taghavi, Seyed Mohammad. 2025. « AI-driven prediction of central plug morphology in Poiseuille flow of Bingham fluids with superhydrophobic walls ». In Proceedings of the CSME-CFDSC-CSR 2025 International Congress (Montreal, QC, Canada, May 25-28, 2025) Coll. « Progress in Canadian Mechanical Engineering », vol. 8.

[thumbnail of 107 - AI-driven prediction of central pl.pdf]
Prévisualisation
PDF
107 - AI-driven prediction of central pl.pdf - Version publiée
Licence d'utilisation : Tous les droits réservés aux détenteurs du droit d'auteur.

Télécharger (729kB) | Prévisualisation

Résumé

This study presents a machine learning-based framework for predicting the flow behavior and central plug morphology of a viscoplastic Poiseuille flow in channels with two superhydrophobic (SH) walls. Here, the lower wall has constant hydrophobicity and groove characteristics, and the upper wall features varying groove periodicity length, slip area fraction, and slip number. Numerical simulations, conducted using OpenFOAM with the Papanastasiou regularization for modelling the viscoplastic fluid, generated a comprehensive database encompassing key flow parameters, including the Bingham number. Four predictive models i.e., Adaptive Neuro-Fuzzy Inference System (ANFIS), Extreme Learning Machine (ELM), Support Vector Machine (SVM), and Multiple Linear Regression (MLR), were used. To assess the accuracy of the models, statistical indices such as correlation coefficient (R), variance accounted for (VAF), root mean square error (RMSE), mean absolute error (MAE), and mean absolute relative error (MARE), were employed. ANFIS demonstrated the highest accuracy, while ELM provided competitive performance with significantly faster computation and simpler hyperparameter tuning. A sensitivity analysis conducted using ELM revealed that the accurate prediction of the normalized area of the center plug is primarily influenced by the groove periodicity length of the upper wall.

Type de document: Compte rendu de conférence
Éditeurs:
Éditeurs
ORCID
Hof, Lucas A.
NON SPÉCIFIÉ
Di Labbio, Giuseppe
NON SPÉCIFIÉ
Tahan, Antoine
NON SPÉCIFIÉ
Sanjosé, Marlène
NON SPÉCIFIÉ
Lalonde, Sébastien
NON SPÉCIFIÉ
Demarquette, Nicole R.
NON SPÉCIFIÉ
Date de dépôt: 18 déc. 2025 15:12
Dernière modification: 18 déc. 2025 15:12
URI: https://espace2.etsmtl.ca/id/eprint/32430

Actions (Authentification requise)

Dernière vérification avant le dépôt Dernière vérification avant le dépôt