A new method based on artificial neural network for radiative heat transfer calculation : Comparison with benchmark numerical solutions in homogeneous media - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

A new method based on artificial neural network for radiative heat transfer calculation : Comparison with benchmark numerical solutions in homogeneous media

Résumé

A new method based on predictive capacity of feedforward artificial neural networks (FANN) is proposed, to estimate the divergence of the radiative flux in a axisymmetric domain very efficiently. Training and validation databases have been built thanks to results given by the SNB-CK model and computed accordingly with Monte Carlo formulation. The major aim of this work is to combine advantages of spectral models in terms of accuracy and the computational efficiency of neural networks and obviously, make possible the precise modeling of radiative transfer on an industrial scale. We show that ANNs are able to model almost perfectly the radiative flux divergence under certain conditions on the basis of training data. We also give here some keys to avoid the pitfalls related to ANNs.
Fichier non déposé

Dates et versions

hal-04107098 , version 1 (26-05-2023)

Identifiants

Citer

Alex Royer, Olivier Farges, Pascal Boulet, Daria Burot. A new method based on artificial neural network for radiative heat transfer calculation : Comparison with benchmark numerical solutions in homogeneous media. Proceedings of CHT-21 ICHMT International Symposium on Advances in Computational Heat Transfer, Aug 2021, Rio de Janeiro, Brazil. pp.271-283, ⟨10.1615/ICHMT.2021.CHT-21.220⟩. ⟨hal-04107098⟩
31 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More