Augmenting mesh-based data-driven models with physics gradients
Augmenting mesh-based data-driven models with physics gradients
Deep learning technologies are increasingly used in various applications, with significant potential in aerospace for reduced-order modelling due to their ability to handle nonlinear systems. The effectiveness of data-driven methods relies on the adequacy and volume of training data, which poses a challenge in a design environment. To address this, physics-informed machine learning, which integrates physics knowledge into data-driven frameworks, has emerged as a promising solution. Directly applying physics terms to aircraft surfaces is complex, so this study utilizes solution gradients to effectively capture flow features. We introduce a hybrid framework that combines geometric deep learning with gradient terms, building on a previous data-driven approach for aerodynamic modelling on large-scale, three-dimensional unstructured grids. We evaluated various hybrid schemes to enhance prediction accuracy. Two gradient-enhanced approaches were found to outperform the purely data-driven model: the first integrates output differentiation into the training loss, achieving the highest accuracy at an increased training cost; the second employs a masking technique to weight regions with large gradients, providing a reasonable accuracy improvement at a lower training cost. This study focuses on predicting distributed aerodynamic loads around the NASA Common Research Model wing/body configuration under various transonic flight conditions. Our findings show that incorporating gradient information into deep learning models significantly improves the accuracy of the predictions and can compensate for a smaller dataset without compromising accuracy. Furthermore, the approaches proposed herein are directly applicable to any problem with discretised spatial domain.
Autoencoder, Common research model, Computational fluid dynamics, Geometric deep learning, Gradient guided, Graph convolutional network, Hybrid approach, Multi mesh, Physics informed, Transonic aerodynamics
Massegur Sampietro, David
d5bc71e8-f1b8-4c9f-9537-7ff63ad19426
Da Ronch, Andrea
a2f36b97-b881-44e9-8a78-dd76fdf82f1a
13 February 2025
Massegur Sampietro, David
d5bc71e8-f1b8-4c9f-9537-7ff63ad19426
Da Ronch, Andrea
a2f36b97-b881-44e9-8a78-dd76fdf82f1a
Massegur Sampietro, David and Da Ronch, Andrea
(2025)
Augmenting mesh-based data-driven models with physics gradients.
Aerospace Science and Technology, 160, [110037].
(doi:10.1016/j.ast.2025.110037).
Abstract
Deep learning technologies are increasingly used in various applications, with significant potential in aerospace for reduced-order modelling due to their ability to handle nonlinear systems. The effectiveness of data-driven methods relies on the adequacy and volume of training data, which poses a challenge in a design environment. To address this, physics-informed machine learning, which integrates physics knowledge into data-driven frameworks, has emerged as a promising solution. Directly applying physics terms to aircraft surfaces is complex, so this study utilizes solution gradients to effectively capture flow features. We introduce a hybrid framework that combines geometric deep learning with gradient terms, building on a previous data-driven approach for aerodynamic modelling on large-scale, three-dimensional unstructured grids. We evaluated various hybrid schemes to enhance prediction accuracy. Two gradient-enhanced approaches were found to outperform the purely data-driven model: the first integrates output differentiation into the training loss, achieving the highest accuracy at an increased training cost; the second employs a masking technique to weight regions with large gradients, providing a reasonable accuracy improvement at a lower training cost. This study focuses on predicting distributed aerodynamic loads around the NASA Common Research Model wing/body configuration under various transonic flight conditions. Our findings show that incorporating gradient information into deep learning models significantly improves the accuracy of the predictions and can compensate for a smaller dataset without compromising accuracy. Furthermore, the approaches proposed herein are directly applicable to any problem with discretised spatial domain.
Text
1-s2.0-S1270963825001087-main
- Version of Record
More information
Accepted/In Press date: 6 February 2025
e-pub ahead of print date: 10 February 2025
Published date: 13 February 2025
Keywords:
Autoencoder, Common research model, Computational fluid dynamics, Geometric deep learning, Gradient guided, Graph convolutional network, Hybrid approach, Multi mesh, Physics informed, Transonic aerodynamics
Identifiers
Local EPrints ID: 501484
URI: http://eprints.soton.ac.uk/id/eprint/501484
ISSN: 1270-9638
PURE UUID: 9c0530c2-0b21-4852-9483-674bbf3172f7
Catalogue record
Date deposited: 02 Jun 2025 16:56
Last modified: 07 Mar 2026 03:22
Export record
Altmetrics
Contributors
Author:
David Massegur Sampietro
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics