 |
 |
Laboratory of Climate and Environmental Sciences (LSCE) - UMR 8212 |  |
 |
Journal articles
Physical invariance in neural networks for subgrid-scale scalar flux modeling
Abstract : In this paper we present a new strategy to model the subgrid-scale scalar flux in a three-dimensional turbulent incompressible flow using physics-informed neural networks (NNs). When trained from direct numerical simulation (DNS) data, state-of-the-art neural networks, such as convolutional neural networks, may not preserve well known physical priors, which may in turn question their application to real case-studies. To address this issue, we investigate hard and soft constraints into the model based on classical invariances and symmetries derived from physical laws. From simulation-based experiments, we show that the proposed physically-invariant NN model outperforms both purely data-driven ones as well as parametric state-of-the-art subgrid-scale model. The considered invariances are regarded as regularizers on physical metrics during the a priori evaluation and constrain the distribution tails of the predicted subgrid-scale term to be closer to the DNS. They also increase the stability and performance of the model when used as a surrogate during a large-eddy simulation. Moreover, the physically-invariant NN is shown to generalize to configurations that have not been seen during the training phase.
https://hal.archives-ouvertes.fr/hal-03084215
Contributor : Hugo Frezat Connect in order to contact the contributor
Submitted on : Tuesday, January 11, 2022 - 5:35:44 PM
Last modification on : Thursday, April 7, 2022 - 1:58:15 PM
Contributor : Hugo Frezat Connect in order to contact the contributor
Submitted on : Tuesday, January 11, 2022 - 5:35:44 PM
Last modification on : Thursday, April 7, 2022 - 1:58:15 PM
File
2010.04663.pdf
Files produced by the author(s)