Data-driven deconvolution for large eddy simulations of Kraichnan turbulence
Abstract
In this article, we demonstrate the use of artificial neural networks as optimal maps which are utilized for convolution and deconvolution of coarse-grained fields to account for sub-grid scale turbulence effects. Here, we demonstrate that an effective eddy-viscosity is predicted by our purely data-driven large eddy simulation framework without explicit utilization of phenomenological arguments. In addition, our data-driven framework precludes the knowledge of true sub-grid stress information during the training phase due to its focus on estimating an effective filter and its inverse so that grid-resolved variables may be related to direct numerical simulation data statistically. The proposed predictive framework is also combined with a statistical truncation mechanism for ensuring numerical realizability in an explicit formulation. Through this, we seek to unite structural and functional modeling strategies for modeling non-linear partial differential equations using reduced degrees of freedom. Both a priori and a posteriori results are shown for a two-dimensional decaying turbulence case in addition to a detailed description of validation and testing. A hyperparameter sensitivity study also shows that the proposed dual network framework simplifies learning complexity and is viable with exceedingly simple network architectures. Our findings indicate that the proposed framework approximates a robust and stable sub-grid closuremore »
- Authors:
-
- Oklahoma State Univ., Stillwater, OK (United States)
- SINTEF Digital, Trondheim (Norway)
- Univ. of Oklahoma, Norman, OK (United States)
- Publication Date:
- Research Org.:
- Oklahoma State Univ., Stillwater, OK (United States)
- Sponsoring Org.:
- USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR)
- OSTI Identifier:
- 1593571
- Grant/Contract Number:
- SC0019290
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Physics of Fluids
- Additional Journal Information:
- Journal Volume: 30; Journal Issue: 12; Journal ID: ISSN 1070-6631
- Publisher:
- American Institute of Physics (AIP)
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 42 ENGINEERING
Citation Formats
Maulik, R., San, O., Rasheed, A., and Vedula, P. Data-driven deconvolution for large eddy simulations of Kraichnan turbulence. United States: N. p., 2018.
Web. doi:10.1063/1.5079582.
Maulik, R., San, O., Rasheed, A., & Vedula, P. Data-driven deconvolution for large eddy simulations of Kraichnan turbulence. United States. https://doi.org/10.1063/1.5079582
Maulik, R., San, O., Rasheed, A., and Vedula, P. Fri .
"Data-driven deconvolution for large eddy simulations of Kraichnan turbulence". United States. https://doi.org/10.1063/1.5079582. https://www.osti.gov/servlets/purl/1593571.
@article{osti_1593571,
title = {Data-driven deconvolution for large eddy simulations of Kraichnan turbulence},
author = {Maulik, R. and San, O. and Rasheed, A. and Vedula, P.},
abstractNote = {In this article, we demonstrate the use of artificial neural networks as optimal maps which are utilized for convolution and deconvolution of coarse-grained fields to account for sub-grid scale turbulence effects. Here, we demonstrate that an effective eddy-viscosity is predicted by our purely data-driven large eddy simulation framework without explicit utilization of phenomenological arguments. In addition, our data-driven framework precludes the knowledge of true sub-grid stress information during the training phase due to its focus on estimating an effective filter and its inverse so that grid-resolved variables may be related to direct numerical simulation data statistically. The proposed predictive framework is also combined with a statistical truncation mechanism for ensuring numerical realizability in an explicit formulation. Through this, we seek to unite structural and functional modeling strategies for modeling non-linear partial differential equations using reduced degrees of freedom. Both a priori and a posteriori results are shown for a two-dimensional decaying turbulence case in addition to a detailed description of validation and testing. A hyperparameter sensitivity study also shows that the proposed dual network framework simplifies learning complexity and is viable with exceedingly simple network architectures. Our findings indicate that the proposed framework approximates a robust and stable sub-grid closure which compares favorably to the Smagorinsky and Leith hypotheses for capturing the theoretical k-3 scaling in Kraichnan turbulence.},
doi = {10.1063/1.5079582},
journal = {Physics of Fluids},
number = 12,
volume = 30,
place = {United States},
year = {Fri Dec 28 00:00:00 EST 2018},
month = {Fri Dec 28 00:00:00 EST 2018}
}
Web of Science
Works referenced in this record:
Diffusion Approximation for Two-Dimensional Turbulence
journal, January 1968
- Leith, C. E.
- Physics of Fluids, Vol. 11, Issue 3
Data-assisted reduced-order modeling of extreme events in complex dynamical systems
journal, May 2018
- Wan, Zhong Yi; Vlachas, Pantelis; Koumoutsakos, Petros
- PLOS ONE, Vol. 13, Issue 5
A survey of decision tree classifier methodology
journal, January 1991
- Safavian, S. R.; Landgrebe, D.
- IEEE Transactions on Systems, Man, and Cybernetics, Vol. 21, Issue 3
Distilling Free-Form Natural Laws from Experimental Data
journal, April 2009
- Schmidt, Michael; Lipson, Hod
- Science, Vol. 324, Issue 5923
A neural network approach for the blind deconvolution of turbulent flows
journal, October 2017
- Maulik, R.; San, O.
- Journal of Fluid Mechanics, Vol. 831
A posteriori analysis of low-pass spatial filters for approximate deconvolution large eddy simulations of homogeneous incompressible flows
journal, December 2014
- San, O.; Staples, A. E.; Iliescu, T.
- International Journal of Computational Fluid Dynamics, Vol. 29, Issue 1
Hybrid forecasting of chaotic processes: Using machine learning in conjunction with a knowledge-based model
journal, April 2018
- Pathak, Jaideep; Wikner, Alexander; Fussell, Rebeckah
- Chaos: An Interdisciplinary Journal of Nonlinear Science, Vol. 28, Issue 4
Determination of the Smagorinsky–Lilly constant CS
journal, May 1997
- Canuto, V. M.; Cheng, Y.
- Physics of Fluids, Vol. 9, Issue 5
Subgrid-scale modeling for implicit large eddy simulation of compressible flows and shock-turbulence interaction
journal, October 2014
- Hickel, Stefan; Egerer, Christian P.; Larsson, Johan
- Physics of Fluids, Vol. 26, Issue 10
Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier–Stokes simulations: A data-driven, physics-informed Bayesian approach
journal, November 2016
- Xiao, H.; Wu, J. -L.; Wang, J. -X.
- Journal of Computational Physics, Vol. 324
A dynamic subgrid‐scale eddy viscosity model
journal, July 1991
- Germano, Massimo; Piomelli, Ugo; Moin, Parviz
- Physics of Fluids A: Fluid Dynamics, Vol. 3, Issue 7
Turbulence Modeling in the Age of Data
journal, January 2019
- Duraisamy, Karthik; Iaccarino, Gianluca; Xiao, Heng
- Annual Review of Fluid Mechanics, Vol. 51, Issue 1
Neural networks based subgrid scale modeling in large eddy simulations
journal, January 2003
- Sarghini, F.; de Felice, G.; Santini, S.
- Computers & Fluids, Vol. 32, Issue 1
Smagorinsky constant in LES modeling of anisotropic MHD turbulence
journal, September 2007
- Vorobev, Anatoliy; Zikanov, Oleg
- Theoretical and Computational Fluid Dynamics, Vol. 22, Issue 3-4
High-order methods for decaying two-dimensional homogeneous isotropic turbulence
journal, June 2012
- San, Omer; Staples, Anne E.
- Computers & Fluids, Vol. 63
Subgrid-scale scalar flux modelling based on optimal estimation theory and machine-learning procedures
journal, June 2017
- Vollant, A.; Balarac, G.; Corre, C.
- Journal of Turbulence, Vol. 18, Issue 9
Evaluation of scale-aware subgrid mesoscale eddy models in a global eddy-rich model
journal, July 2017
- Pearson, Brodie; Fox-Kemper, Baylor; Bachman, Scott
- Ocean Modelling, Vol. 115
Deep Learning in Medical Image Analysis
journal, June 2017
- Shen, Dinggang; Wu, Guorong; Suk, Heung-Il
- Annual Review of Biomedical Engineering, Vol. 19, Issue 1
A framework for large eddy simulation of Burgers turbulence based upon spatial and temporal statistical information
journal, March 2015
- LaBryer, A.; Attar, P. J.; Vedula, P.
- Physics of Fluids, Vol. 27, Issue 3
Approximate deconvolution model for the simulation of turbulent gas-solid flows: An a priori analysis
journal, February 2018
- Schneiderbauer, Simon; Saeedipour, Mahdi
- Physics of Fluids, Vol. 30, Issue 2
Deep Learning and Its Application to LHC Physics
journal, October 2018
- Guest, Dan; Cranmer, Kyle; Whiteson, Daniel
- Annual Review of Nuclear and Particle Science, Vol. 68, Issue 1
Learning partial differential equations via data discovery and sparse optimization
journal, January 2017
- Schaeffer, Hayden
- Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 473, Issue 2197
The development of algebraic stress models using a novel evolutionary algorithm
journal, December 2017
- Weatheritt, J.; Sandberg, R. D.
- International Journal of Heat and Fluid Flow, Vol. 68
Neural network closures for nonlinear model order reduction
journal, January 2018
- San, Omer; Maulik, Romit
- Advances in Computational Mathematics, Vol. 44, Issue 6
A stable and scale-aware dynamic modeling framework for subgrid-scale parameterizations of two-dimensional turbulence
journal, November 2017
- Maulik, Romit; San, Omer
- Computers & Fluids, Vol. 158
Using field inversion to quantify functional errors in turbulence closures
journal, April 2016
- Singh, Anand Pratap; Duraisamy, Karthik
- Physics of Fluids, Vol. 28, Issue 4
Using statistical learning to close two-fluid multiphase flow equations for a simple bubbly system
journal, September 2015
- Ma, Ming; Lu, Jiacai; Tryggvason, Gretar
- Physics of Fluids, Vol. 27, Issue 9
Prediction model of velocity field around circular cylinder over various Reynolds numbers by fusion convolutional neural networks based on pressure on the cylinder
journal, April 2018
- Jin, Xiaowei; Cheng, Peng; Chen, Wen-Li
- Physics of Fluids, Vol. 30, Issue 4
Multilayer feedforward networks are universal approximators
journal, January 1989
- Hornik, Kurt; Stinchcombe, Maxwell; White, Halbert
- Neural Networks, Vol. 2, Issue 5
Machine learning of linear differential equations using Gaussian processes
journal, November 2017
- Raissi, Maziar; Perdikaris, Paris; Karniadakis, George Em
- Journal of Computational Physics, Vol. 348
Hidden physics models: Machine learning of nonlinear partial differential equations
journal, March 2018
- Raissi, Maziar; Karniadakis, George Em
- Journal of Computational Physics, Vol. 357
Subgrid modelling for two-dimensional turbulence using neural networks
journal, November 2018
- Maulik, R.; San, O.; Rasheed, A.
- Journal of Fluid Mechanics, Vol. 858
Hybrid Reynolds-Averaged/Large-Eddy Simulation Methodology from Symbolic Regression: Formulation and Application
journal, November 2017
- Weatheritt, Jack; Sandberg, Richard D.
- AIAA Journal, Vol. 55, Issue 11
Application of machine learning to viscoplastic flow modeling
journal, October 2018
- Muravleva, E.; Oseledets, I.; Koroteev, D.
- Physics of Fluids, Vol. 30, Issue 10
Mathematical Perspectives on Large Eddy Simulation Models for Turbulent Flows
journal, June 2004
- Guermond, J. -L.; Oden, J. T.; Prudhomme, S.
- Journal of Mathematical Fluid Mechanics, Vol. 6, Issue 2
Discovering governing equations from data by sparse identification of nonlinear dynamical systems
journal, March 2016
- Brunton, Steven L.; Proctor, Joshua L.; Kutz, J. Nathan
- Proceedings of the National Academy of Sciences, Vol. 113, Issue 15
Reynolds averaged turbulence modelling using deep neural networks with embedded invariance
journal, October 2016
- Ling, Julia; Kurzawski, Andrew; Templeton, Jeremy
- Journal of Fluid Mechanics, Vol. 807
Inertial Ranges in Two-Dimensional Turbulence
journal, January 1967
- Kraichnan, Robert H.
- Physics of Fluids, Vol. 10, Issue 7
Dynamic optimization methodology based on subgrid-scale dissipation for large eddy simulation
journal, January 2016
- Yu, Changping; Xiao, Zuoli; Li, Xinliang
- Physics of Fluids, Vol. 28, Issue 1
A novel evolutionary algorithm applied to algebraic modifications of the RANS stress–strain relationship
journal, November 2016
- Weatheritt, Jack; Sandberg, Richard
- Journal of Computational Physics, Vol. 325
Neural Network Modeling for Near Wall Turbulent Flow
journal, October 2002
- Milano, Michele; Koumoutsakos, Petros
- Journal of Computational Physics, Vol. 182, Issue 1
Compressive sensing based machine learning strategy for characterizing the flow around a cylinder with limited pressure measurements
journal, December 2013
- Bright, Ido; Lin, Guang; Kutz, J. Nathan
- Physics of Fluids, Vol. 25, Issue 12
Evaluation of machine learning algorithms for prediction of regions of high Reynolds averaged Navier Stokes uncertainty
journal, August 2015
- Ling, J.; Templeton, J.
- Physics of Fluids, Vol. 27, Issue 8
Theoretically based optimal large-eddy simulation
journal, October 2009
- Moser, Robert D.; Malaya, Nicholas P.; Chang, Henry
- Physics of Fluids, Vol. 21, Issue 10
An approximate deconvolution procedure for large-eddy simulation
journal, July 1999
- Stolz, S.; Adams, N. A.
- Physics of Fluids, Vol. 11, Issue 7
Non-intrusive reduced-order modelling of the Navier-Stokes equations based on RBF interpolation: Non-intrusive reduced-order modelling of the Navier-Stokes equations based on RBF interpolation
journal, July 2015
- Xiao, D.; Fang, F.; Pain, C.
- International Journal for Numerical Methods in Fluids, Vol. 79, Issue 11
Machine Learning
journal, June 1990
- Mitchell, T.; Buchanan, B.; DeJong, G.
- Annual Review of Computer Science, Vol. 4, Issue 1
The similarity subgrid stresses associated to the approximate Van Cittert deconvolutions
journal, March 2015
- Germano, M.
- Physics of Fluids, Vol. 27, Issue 3
Closed-loop separation control using machine learning
journal, April 2015
- Gautier, N.; Aider, J. -L.; Duriez, T.
- Journal of Fluid Mechanics, Vol. 770
Deep learning in fluid dynamics
journal, January 2017
- Kutz, J. Nathan
- Journal of Fluid Mechanics, Vol. 814
Analysis and development of subgrid turbulence models preserving the symmetry properties of the Navier–Stokes equations
journal, July 2007
- Razafindralandy, Dina; Hamdouni, Aziz; Oberlack, Martin
- European Journal of Mechanics - B/Fluids, Vol. 26, Issue 4
Works referencing / citing this record:
Sub-grid scale model classification and blending through deep learning
journal, May 2019
- Maulik, Romit; San, Omer; Jacob, Jamey D.
- Journal of Fluid Mechanics, Vol. 870
Prediction of turbulent heat transfer using convolutional neural networks
journal, November 2019
- Kim, Junhyuk; Lee, Changhoon
- Journal of Fluid Mechanics, Vol. 882
Fast flow field prediction over airfoils using deep learning approach
journal, May 2019
- Sekar, Vinothkumar; Khoo, Boo Cheong
- Physics of Fluids, Vol. 31, Issue 5
Linking dissipation, anisotropy, and intermittency in rotating stratified turbulence at the threshold of linear shear instabilities
journal, October 2019
- Pouquet, A.; Rosenberg, D.; Marino, R.
- Physics of Fluids, Vol. 31, Issue 10
Memory embedded non-intrusive reduced order modeling of non-ergodic flows
journal, December 2019
- Ahmed, Shady E.; Rahman, Sk. Mashfiqur; San, Omer
- Physics of Fluids, Vol. 31, Issue 12
Feature engineering and symbolic regression methods for detecting hidden physics from sparse sensor observation data
journal, January 2020
- Vaddireddy, Harsha; Rasheed, Adil; Staples, Anne E.
- Physics of Fluids, Vol. 32, Issue 1
Sub-grid scale model classification and blending through deep learning
text, January 2018
- Maulik, Romit; San, Omer; Jacob, Jamey D.
- arXiv
Memory embedded non-intrusive reduced order modeling of non-ergodic flows
text, January 2019
- Ahmed, Shady E.; Rahman, Sk. Mashfiqur; San, Omer
- arXiv
Review of Physics-based and Data-driven Multiscale Simulation Methods for Computational Fluid Dynamics and Nuclear Thermal Hydraulics
preprint, January 2021
- Iskhakov, Arsen S.; Dinh, Nam T.
- arXiv
Investigation of nonlocal data-driven methods for subgrid-scale stress modelling in large eddy simulation
text, January 2021
- Liu, Bo; Yu, Huiyang; Huang, Haibo
- arXiv