Physics-enhanced neural networks for equation-of-state calculations

Timothy J. Callow*, Jan Nikl, Eli Kraisler, Attila Cangi

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Rapid access to accurate equation-of-state (EOS) data is crucial in the warm-dense matter (WDM) regime, as it is employed in various applications, such as providing input for hydrodynamic codes to model inertial confinement fusion processes. In this study, we develop neural network models for predicting the EOS based on first-principles data. The first model utilises basic physical properties, while the second model incorporates more sophisticated physical information, using output from average-atom (AA) calculations as features. AA models are often noted for providing a reasonable balance of accuracy and speed; however, our comparison of AA models and higher-fidelity calculations shows that more accurate models are required in the WDM regime. Both the neural network models we propose, particularly the physics-enhanced one, demonstrate significant potential as accurate and efficient methods for computing EOS data in WDM.

Original languageAmerican English
Article number045055
JournalMachine Learning: Science and Technology
Volume4
Issue number4
DOIs
StatePublished - 1 Dec 2023

Bibliographical note

Publisher Copyright:
© 2023 The Author(s). Published by IOP Publishing Ltd.

Keywords

  • average-atom models
  • equation-of-state calculations
  • physics-enhanced neural networks
  • warm dense matter

Fingerprint

Dive into the research topics of 'Physics-enhanced neural networks for equation-of-state calculations'. Together they form a unique fingerprint.

Cite this