Neural network-based modelling of unresolved stresses in a turbulent reacting flow with mean shear

Data-driven methods for modelling purposes in fluid mechanics are a promising alternative given the continuous increase of both computational power and data-storage capabilities. Highly non-linear flows including turbulence and reaction are challenging to model, and accurate closures for the unresolved terms in large eddy simulations of such flows are difficult to obtain. In this study, we investigate the use of artificial neural networks for modelling an important unclosed term namely the unresolved stress tensor, in a highly demanding turbulent and reacting flow, which additionally includes mean shear. The performance of the neural network-based modelling approach is conducted a priori following a coarsened mesh approach, and compared against the predictions of eight other classic models in the literature, which include both static and dynamic formulations.

[1]  Jacqueline H. Chen,et al.  Direct numerical simulations of a high Karlovitz number laboratory premixed jet flame – an analysis of flame stretch and flame thickening , 2017, Journal of Fluid Mechanics.

[2]  F. Sarghini,et al.  Neural networks based subgrid scale modeling in large eddy simulations , 2003 .

[3]  G.,et al.  TOWARD THE LARGE-EDDY SIMULATION OF COMPRESSIBLE TURBULENT FLOWS , 2022 .

[4]  Robert J. Kee,et al.  A FORTRAN COMPUTER CODE PACKAGE FOR THE EVALUATION OF GAS-PHASE, MULTICOMPONENT TRANSPORT PROPERTIES , 1986 .

[5]  T. Poinsot,et al.  Training convolutional neural networks to estimate turbulent sub-grid scale reaction rates , 2018, Combustion and Flame.

[6]  Karthik Duraisamy,et al.  Machine Learning-augmented Predictive Modeling of Turbulent Separated Flows over Airfoils , 2016, ArXiv.

[7]  D. Lilly,et al.  A proposed modification of the Germano subgrid‐scale closure method , 1992 .

[8]  J. Chomiak,et al.  Modeling variable density effects in turbulent flames—Some basic considerations , 1995 .

[9]  F. Nicoud,et al.  Subgrid-Scale Stress Modelling Based on the Square of the Velocity Gradient Tensor , 1999 .

[10]  Baris A. Sen,et al.  Large eddy simulation of extinction and reignition with artificial neural networks based chemical kinetics , 2010 .

[11]  J. B. Moss,et al.  Turbulence Production in Premixed Turbulent Flames , 1981 .

[12]  P. Moin,et al.  A dynamic subgrid‐scale model for compressible turbulence and scalar transport , 1991 .

[13]  Hervé Jeanmart,et al.  Explicit-filtering large-eddy simulation using the tensor-diffusivity model supplemented by a dynami , 2001 .

[14]  M. Klein,et al.  A-priori direct numerical simulation assessment of sub-grid scale stress tensor closures for turbulent premixed combustion , 2015 .

[15]  Yuji Hattori,et al.  Searching for turbulence models by artificial neural network , 2016, 1607.01042.

[16]  Bok Jik Lee,et al.  Direct numerical simulations of reacting flows with detailed chemistry using many-core/GPU acceleration , 2018, Computers & Fluids.

[17]  Michael Bauer,et al.  S3D-Legion : An Exascale Software for Direct Numerical Simulation of Turbulent Combustion with Complex Multicomponent Chemistry , 2017 .

[18]  Rs Cant,et al.  Scalar flux modeling in turbulent flames using iterative deconvolution , 2018 .

[19]  Hans Kuerten,et al.  Large-eddy simulation of the temporal mixing layer using the Clark model , 1996 .

[20]  Jinlong Wu,et al.  Physics-informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data , 2016, 1606.07987.

[21]  Kurt Hornik,et al.  Approximation capabilities of multilayer feedforward networks , 1991, Neural Networks.

[22]  J. Nathan Kutz,et al.  Deep learning in fluid dynamics , 2017, Journal of Fluid Mechanics.

[23]  H. Pitsch.  LARGE-EDDY SIMULATION OF TURBULENT COMBUSTION , 2006 .

[24]  Thierry Poinsot,et al.  Large Eddy Simulations of gaseous flames in gas turbine combustion chambers , 2012 .

[25]  Baris A. Sen,et al.  Turbulent premixed flame modeling using artificial neural networks based chemical kinetics , 2009 .

[26]  J. Ferziger,et al.  Improved turbulence models based on large eddy simulation of homogeneous, incompressible, turbulent flows , 1983 .

[27]  C. Meneveau,et al.  Scale-Invariance and Turbulence Models for Large-Eddy Simulation , 2000 .

[28]  Randal C. Burns,et al.  The Johns Hopkins Turbulence Databases: An Open Simulation Laboratory for Turbulence Research , 2015, Computing in Science & Engineering.

[29]  Demis Hassabis,et al.  Mastering the game of Go with deep neural networks and tree search , 2016, Nature.

[30]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[31]  Charalambos Chrysostomou,et al.  Progress Variable Variance and Filtered Rate Modelling Using Convolutional Neural Networks and Flamelet Methods , 2019, Flow, Turbulence and Combustion.

[32]  A. Leipertz,et al.  A-priori testing of an eddy viscosity model for the density-weighted subgrid scale stress tensor in turbulent premixed flames , 2010 .

[33]  C. Meneveau,et al.  Geometry and scale relationships in high Reynolds number turbulence determined from three-dimensional holographic velocimetry , 2000 .

[34]  Effect of flow-geometry on turbulence-scalar interaction in premixed flames , 2011 .

[35]  Maria Vittoria Salvetti,et al.  A Priori Tests of a New Dynamic Subgrid-Scale Model for Finite-Difference Large-Eddy Simulations , 1995 .

[36]  Luc Vervisch,et al.  A compressible wall-adapting similarity mixed model for large-eddy simulation of the impinging round jet , 2009 .

[37]  J. Smagorinsky,et al.  GENERAL CIRCULATION EXPERIMENTS WITH THE PRIMITIVE EQUATIONS , 1963 .

[38]  Markus Klein,et al.  Application of an evolutionary algorithm to LES modelling of turbulent transport in premixed flames , 2018, J. Comput. Phys..

[39]  M. Ringnér,et al.  Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks , 2001, Nature Medicine.

[40]  Marcus S. Day,et al.  Three-dimensional direct numerical simulation of turbulent lean premixed methane combustion with detailed kinetics , 2016 .

[41]  J. Templeton,et al.  Reynolds averaged turbulence modelling using deep neural networks with embedded invariance , 2016, Journal of Fluid Mechanics.

[42]  Omer San,et al.  A neural network approach for the blind deconvolution of turbulent flows , 2017, Journal of Fluid Mechanics.

[43]  Nedunchezhian Swaminathan,et al.  Direct Numerical Simulation of Complex Fuel Combustion with Detailed Chemistry: Physical Insight and Mean Reaction Rate Modeling , 2015 .

[44]  Michele Milano,et al.  Neural network modeling for near wall turbulent flow , 2002 .

[45]  Brendan D. Tracey,et al.  A Machine Learning Strategy to Assist Turbulence Model Development , 2015 .

[46]  Parviz Moin,et al.  Erratum: ‘‘A dynamic subgrid‐scale eddy viscosity model’’ [Phys. Fluids A 3, 1760 (1991)] , 1991 .

[47]  J. Andrzej Domaradzki,et al.  A subgrid-scale model for large-eddy simulation based on the physics of interscale energy transfer in turbulence , 2012 .

[48]  F. A. Williams,et al.  Structure and Extinction of Hydrogen-Air Diffusion Flames , 1993 .

[49]  Charles Meneveau,et al.  Statistical geometry of subgrid-scale stresses determined from holographic particle image velocimetry measurements , 2002, Journal of Fluid Mechanics.

[50]  L. Vervisch,et al.  Modelling turbulent premixed flames using convolutional neural networks: application to sub-grid scale variance and filtered reaction rate. , 2018, 1810.07944.

[51]  J. Koseff,et al.  A dynamic mixed subgrid‐scale model and its application to turbulent recirculating flows , 1993 .

[52]  J. Ferziger,et al.  Evaluation of subgrid-scale models using an accurately simulated turbulent flow , 1979, Journal of Fluid Mechanics.

[53]  Shane Legg,et al.  Human-level control through deep reinforcement learning , 2015, Nature.

[54]  Martín Abadi,et al.  TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems , 2016, ArXiv.

[55]  Nedunchezhian Swaminathan,et al.  Evaluation of a reduced mechanism for turbulent premixed combustion , 2014 .

[56]  Jianren Fan,et al.  Investigations of data-driven closure for subgrid-scale stress in large-eddy simulation , 2018, Physics of Fluids.

[57]  Quoc V. Le,et al.  Sequence to Sequence Learning with Neural Networks , 2014, NIPS.

[58]  Christoph Schmitt,et al.  Optimal artificial neural networks and tabulation methods for chemistry representation in LES of a bluff-body swirl-stabilized flame , 2009 .

[59]  Olivier Teytaud,et al.  Optimal estimation for Large-Eddy Simulation of turbulence and application to the analysis of subgrid models , 2006, ArXiv.

[60]  A. Yoshizawa.  Statistical theory for compressible turbulent shear flows, with the application to subgrid modeling , 1986 .

[61]  Luc Vervisch,et al.  A Priori Assessment of an Iterative Deconvolution Method for LES Sub-grid Scale Variance Modelling , 2018 .

[62]  G. Tryggvason,et al.  Using statistical learning to close two-fluid multiphase flow equations for a simple bubbly system , 2015 .

[63]  R. J. Kee,et al.  Chemkin-II : A Fortran Chemical Kinetics Package for the Analysis of Gas Phase Chemical Kinetics , 1991 .

[64]  P. Sagaut.  BOOK REVIEW: Large Eddy Simulation for Incompressible Flows. An Introduction , 2001 .