Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations
Abstract
For centuries, flow visualization has been the art of making fluid motion visible in physical and biological systems. Although such flow patterns can be, in principle, described by the Navier-Stokes equations, extracting the velocity and pressure fields directly from the images is challenging. We addressed this problem by developing hidden fluid mechanics (HFM), a physics-informed deep-learning framework capable of encoding the Navier-Stokes equations into the neural networks while being agnostic to the geometry or the initial and boundary conditions. We demonstrate HFM for several physical and biomedical problems by extracting quantitative information for which direct measurements may not be possible.
- Authors:
-
- University of Colorado at Boulder
- Brown University
- BROWN UNIVERSITY
- Publication Date:
- Research Org.:
- Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
- Sponsoring Org.:
- USDOE
- OSTI Identifier:
- 1631426
- Report Number(s):
- PNNL-SA-152906
- DOE Contract Number:
- AC05-76RL01830
- Resource Type:
- Journal Article
- Journal Name:
- Science
- Additional Journal Information:
- Journal Volume: 367; Journal Issue: 6481
- Country of Publication:
- United States
- Language:
- English
Citation Formats
Raissi, Maziar, Yazdani, Alireza, and Karniadakis, George E. Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations. United States: N. p., 2020.
Web. doi:10.1126/science.aaw4741.
Raissi, Maziar, Yazdani, Alireza, & Karniadakis, George E. Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations. United States. https://doi.org/10.1126/science.aaw4741
Raissi, Maziar, Yazdani, Alireza, and Karniadakis, George E. 2020.
"Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations". United States. https://doi.org/10.1126/science.aaw4741.
@article{osti_1631426,
title = {Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations},
author = {Raissi, Maziar and Yazdani, Alireza and Karniadakis, George E.},
abstractNote = {For centuries, flow visualization has been the art of making fluid motion visible in physical and biological systems. Although such flow patterns can be, in principle, described by the Navier-Stokes equations, extracting the velocity and pressure fields directly from the images is challenging. We addressed this problem by developing hidden fluid mechanics (HFM), a physics-informed deep-learning framework capable of encoding the Navier-Stokes equations into the neural networks while being agnostic to the geometry or the initial and boundary conditions. We demonstrate HFM for several physical and biomedical problems by extracting quantitative information for which direct measurements may not be possible.},
doi = {10.1126/science.aaw4741},
url = {https://www.osti.gov/biblio/1631426},
journal = {Science},
number = 6481,
volume = 367,
place = {United States},
year = {Fri Feb 28 00:00:00 EST 2020},
month = {Fri Feb 28 00:00:00 EST 2020}
}
Works referenced in this record:
The effect of a discrete window offset on the accuracy of cross-correlation analysis of digital PIV recordings
journal, May 1997
- Westerweel, J.; Dabiri, D.; Gharib, M.
- Experiments in Fluids, Vol. 23, Issue 1
Microscale 3D flow mapping with μDDPIV
journal, February 2007
- Pereira, Francisco; Lu, Jian; Castaño-Graff, Emilio
- Experiments in Fluids, Vol. 42, Issue 4
Convection and flow in porous media. Part 1. Visualization by magnetic resonance imaging
journal, February 1997
- Shattuck, M. D.; Behringer, R. P.; Johnson, G. A.
- Journal of Fluid Mechanics, Vol. 332
Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
journal, February 2019
- Raissi, M.; Perdikaris, P.; Karniadakis, G. E.
- Journal of Computational Physics, Vol. 378
A convergence study of a new partitioned fluid–structure interaction algorithm based on fictitious mass and damping
journal, January 2012
- Baek, Hyoungsu; Karniadakis, George Em
- Journal of Computational Physics, Vol. 231, Issue 2
Laser-induced fluorescence measurements of mixed fluid concentrationin a liquid plane shear layer
journal, November 1985
- Koochesfahani, M. M.; Dimotakis, P. E.
- AIAA Journal, Vol. 23, Issue 11
Planar laser induced fluorescence in aqueous flows
journal, April 2008
- Crimaldi, J. P.
- Experiments in Fluids, Vol. 44, Issue 6
Coronary Atherosclerosis Imaging by Coronary CT Angiography
journal, May 2011
- Voros, Szilard; Rinehart, Sarah; Qian, Zhen
- JACC: Cardiovascular Imaging, Vol. 4, Issue 5
Prognostic accuracy of cerebral blood flow measurement by perfusion computed tomography, at the time of emergency room admission, in acute stroke patients
journal, March 2002
- Wintermark, Max; Reichhart, Marc; Thiran, Jean-Philippe
- Annals of Neurology, Vol. 51, Issue 4
Triangulating a Cognitive Control Network Using Diffusion-Weighted Magnetic Resonance Imaging (MRI) and Functional MRI
journal, April 2007
- Aron, A. R.; Behrens, T. E.; Smith, S.
- Journal of Neuroscience, Vol. 27, Issue 14
Wall Shear Stress and Early Atherosclerosis: A Review
journal, June 2000
- Shaaban, Akram M.; Duerinckx, André J.
- American Journal of Roentgenology, Vol. 174, Issue 6
Carotid bifurcation atherosclerosis. Quantitative correlation of plaque localization with flow velocity profiles and wall shear stress.
journal, October 1983
- Zarins, C. K.; Giddens, D. P.; Bharadvaj, B. K.
- Circulation Research, Vol. 53, Issue 4
Phase-contrast magnetic resonance imaging measurements in intracranial aneurysms in vivo of flow patterns, velocity fields, and wall shear stress: Comparison with computational fluid dynamics
journal, February 2009
- Boussel, Loic; Rayz, Vitaliy; Martin, Alastair
- Magnetic Resonance in Medicine, Vol. 61, Issue 2
On the Correlation of Parity and Small-Depth Circuits
journal, January 2014
- Håstad, Johan
- SIAM Journal on Computing, Vol. 43, Issue 5
Data-driven discovery of partial differential equations
journal, April 2017
- Rudy, Samuel H.; Brunton, Steven L.; Proctor, Joshua L.
- Science Advances, Vol. 3, Issue 4
Approximation by superpositions of a sigmoidal function
journal, December 1989
- Cybenko, G.
- Mathematics of Control, Signals, and Systems, Vol. 2, Issue 4
A Survey on Transfer Learning
journal, October 2010
- Pan, Sinno Jialin; Yang, Qiang
- IEEE Transactions on Knowledge and Data Engineering, Vol. 22, Issue 10
An Album of Fluid Motion
journal, December 1982
- van Dyke, Milton; White, Frank M.
- Journal of Fluids Engineering, Vol. 104, Issue 4
A low-dimensional model for simulating three-dimensional cylinder flow
journal, May 2002
- Ma, Xia; Karniadakis, George Em
- Journal of Fluid Mechanics, Vol. 458
Three-dimensional dynamics and transition to turbulence in the wake of bluff objects
journal, May 1992
- Karniadakis, George Em; Triantafyllou, George S.
- Journal of Fluid Mechanics, Vol. 238
Inflammation, Atherosclerosis, and Coronary Artery Disease
journal, April 2005
- Hansson, Göran K.
- New England Journal of Medicine, Vol. 352, Issue 16
Outflow Boundary Conditions for Arterial Networks with Multiple Outlets
journal, July 2008
- Grinberg, Leopold; Karniadakis, George Em
- Annals of Biomedical Engineering, Vol. 36, Issue 9