Using physics-informed generative adversarial networks to perform super-resolution for multiphase fluid simulations

Matthew Li, Christopher McComb

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Computational Fluid Dynamics (CFD) simulations are useful to the field of engineering design as they provide deep insights on product or system performance without the need to construct and test physical prototypes. However, they can be very computationally intensive to run. Machine learning methods have been shown to reconstruct high-resolution single-phase turbulent fluid flow simulations from low-resolution inputs. This offers a potential avenue towards alleviating computational cost in iterative engineering design applications. However, little work thus far has explored the application of machine learning image super-resolution methods to multiphase fluid flow (which is important for important for emerging fields such as marine hydrokinetic energy conversion). In this work, we apply a modified version of the Super-Resolution Generative Adversarial Network (SRGAN) model to a multiphase turbulent fluid flow problem, specifically to reconstruct fluid phase fraction at a higher resolution. Two models were created in this work, one with a simple physics-constrained loss function and one without, and the results are discussed and analyzed. We found that both models were able to significantly outperform non-machine learning upsampling methods and can preserve an impressive amount of detail and nuance, showing the versatility of the SRGAN model for upsampling fluid simulations. However, the difference in accuracy between the two models is quite minimal. This indicates that, for these contexts studied here, the additional complexity of a physics-informed approach may not be justified.

Original languageEnglish (US)
Title of host publication47th Design Automation Conference (DAC)
PublisherAmerican Society of Mechanical Engineers (ASME)
ISBN (Electronic)9780791885383
DOIs
StatePublished - 2021
Event47th Design Automation Conference, DAC 2021, Held as Part of the ASME 2021 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, IDETC-CIE 2021 - Virtual, Online
Duration: Aug 17 2021Aug 19 2021

Publication series

NameProceedings of the ASME Design Engineering Technical Conference
Volume3A-2021

Conference

Conference47th Design Automation Conference, DAC 2021, Held as Part of the ASME 2021 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference, IDETC-CIE 2021
CityVirtual, Online
Period8/17/218/19/21

All Science Journal Classification (ASJC) codes

  • Mechanical Engineering
  • Computer Graphics and Computer-Aided Design
  • Computer Science Applications
  • Modeling and Simulation

Cite this