DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Deep learning-based model for progress variable dissipation rate in turbulent premixed flames

Abstract

A deep neural network (DNN) based large eddy simulation (LES) model for progress variable dissipation rate in turbulent premixed flames is presented. The DNN model is trained using filtered data from direct numerical simulations (DNS) of statistically planar turbulent premixed flames with n-heptane as fuel. Training data was comprised of flames with varying turbulence levels leading to a range of Karlovitz numbers. Through a-priori tests the DNN model is shown to predict the subfilter contribution to progress variable dissipation rate accurately over a range of filter widths and for all Karlovitz numbers examined in this study. Superior performance of the DNN model relative to an established physics-based model is also demonstrated. Additionally, transferability of the DNN model is highlighted by a-priori evaluation of the model using filtered DNS data from multiple cases with different Karlovitz numbers and fuel species than those that were used for training the model.

Authors:
ORCiD logo [1];  [1];  [1]
  1. National Renewable Energy Lab. (NREL), Golden, CO (United States)
Publication Date:
Research Org.:
National Renewable Energy Lab. (NREL), Golden, CO (United States)
Sponsoring Org.:
USDOE Office of Energy Efficiency and Renewable Energy (EERE), Vehicle Technologies Office (EE-3V)
OSTI Identifier:
1665878
Report Number(s):
NREL/JA-2C00-75428
Journal ID: ISSN 1540-7489; MainId:6223;UUID:4d7fbbc7-8a05-ea11-9c2a-ac162d87dfe5;MainAdminID:18561
Grant/Contract Number:  
AC36-08GO28308
Resource Type:
Accepted Manuscript
Journal Name:
Proceedings of the Combustion Institute
Additional Journal Information:
Journal Volume: 38; Journal Issue: 2; Journal ID: ISSN 1540-7489
Publisher:
Elsevier
Country of Publication:
United States
Language:
English
Subject:
33 ADVANCED PROPULSION SYSTEMS; deep neural network (DNN); large eddy simulation (LES); premixed flames; progress variable dissipation rate

Citation Formats

Yellapantula, Shashank, Perry, Bruce A., and Grout, Ray W. Deep learning-based model for progress variable dissipation rate in turbulent premixed flames. United States: N. p., 2020. Web. doi:10.1016/j.proci.2020.06.205.
Yellapantula, Shashank, Perry, Bruce A., & Grout, Ray W. Deep learning-based model for progress variable dissipation rate in turbulent premixed flames. United States. https://doi.org/10.1016/j.proci.2020.06.205
Yellapantula, Shashank, Perry, Bruce A., and Grout, Ray W. Fri . "Deep learning-based model for progress variable dissipation rate in turbulent premixed flames". United States. https://doi.org/10.1016/j.proci.2020.06.205. https://www.osti.gov/servlets/purl/1665878.
@article{osti_1665878,
title = {Deep learning-based model for progress variable dissipation rate in turbulent premixed flames},
author = {Yellapantula, Shashank and Perry, Bruce A. and Grout, Ray W.},
abstractNote = {A deep neural network (DNN) based large eddy simulation (LES) model for progress variable dissipation rate in turbulent premixed flames is presented. The DNN model is trained using filtered data from direct numerical simulations (DNS) of statistically planar turbulent premixed flames with n-heptane as fuel. Training data was comprised of flames with varying turbulence levels leading to a range of Karlovitz numbers. Through a-priori tests the DNN model is shown to predict the subfilter contribution to progress variable dissipation rate accurately over a range of filter widths and for all Karlovitz numbers examined in this study. Superior performance of the DNN model relative to an established physics-based model is also demonstrated. Additionally, transferability of the DNN model is highlighted by a-priori evaluation of the model using filtered DNS data from multiple cases with different Karlovitz numbers and fuel species than those that were used for training the model.},
doi = {10.1016/j.proci.2020.06.205},
journal = {Proceedings of the Combustion Institute},
number = 2,
volume = 38,
place = {United States},
year = {Fri Sep 11 00:00:00 EDT 2020},
month = {Fri Sep 11 00:00:00 EDT 2020}
}

Works referenced in this record:

Premixed turbulent combustion modeling using tabulated detailed chemistry and PDF
journal, January 2005


Modified laminar flamelet presumed probability density function for LES of premixed turbulent combustion
journal, January 2013

  • Mahdi Salehi, M.; Kendal Bushe, W.; Shahbazian, Nasim
  • Proceedings of the Combustion Institute, Vol. 34, Issue 1
  • DOI: 10.1016/j.proci.2012.06.177

A priori filtered chemical source term modeling for LES of high Karlovitz number premixed flames
journal, February 2017


Scalar dissipation rate modelling for Large Eddy Simulation of turbulent premixed flames
journal, January 2013

  • Dunstan, T. D.; Minamoto, Y.; Chakraborty, N.
  • Proceedings of the Combustion Institute, Vol. 34, Issue 1
  • DOI: 10.1016/j.proci.2012.06.143

Algebraic Closure of Scalar Dissipation Rate for Large Eddy Simulations of Turbulent Premixed Combustion
journal, September 2014


Turbulent combustion modeling
journal, March 2002


A two mixture fraction flamelet model for large eddy simulation of turbulent flames with inhomogeneous inlets
journal, January 2017

  • Perry, Bruce A.; Mueller, Michael E.; Masri, Assaad R.
  • Proceedings of the Combustion Institute, Vol. 36, Issue 2
  • DOI: 10.1016/j.proci.2016.07.029

Turbulent premixed combustion: Further discussions on the scales of fluctuations
journal, June 1990


Scalar Dissipation Rate Modeling and its Validation
journal, February 2009

  • Kolla, H.; Rogerson, J. W.; Chakraborty, N.
  • Combustion Science and Technology, Vol. 181, Issue 3
  • DOI: 10.1080/00102200802612419

Scalar Dissipation Rate Transport in the Context of Large Eddy Simulations for Turbulent Premixed Flames with Non-Unity Lewis Number
journal, July 2014


Dynamic Closure of Scalar Dissipation Rate for Large Eddy Simulations of Turbulent Premixed Combustion: A Direct Numerical Simulations Analysis
journal, August 2015


Deep learning for presumed probability density function models
journal, October 2019


Training convolutional neural networks to estimate turbulent sub-grid scale reaction rates
journal, May 2019


Differential diffusion effects, distributed burning, and local extinctions in high Karlovitz premixed flames
journal, September 2015


Interaction of turbulence and scalar fields in premixed flames
journal, April 2006

  • Swaminathan, N.; Grout, R. W.
  • Physics of Fluids, Vol. 18, Issue 4
  • DOI: 10.1063/1.2186590

Influence of combustion on principal strain-rate transport in turbulent premixed flames
journal, January 2015

  • Steinberg, A. M.; Coriton, B.; Frank, J. H.
  • Proceedings of the Combustion Institute, Vol. 35, Issue 2
  • DOI: 10.1016/j.proci.2014.06.089

Influence of the Damköhler number on turbulence-scalar interaction in premixed flames. II. Model development
journal, April 2007

  • Chakraborty, N.; Swaminathan, N.
  • Physics of Fluids, Vol. 19, Issue 4
  • DOI: 10.1063/1.2714076

Interactions between turbulence and flames in premixed reacting flows
journal, December 2011

  • Hamlington, Peter E.; Poludnenko, Alexei Y.; Oran, Elaine S.
  • Physics of Fluids, Vol. 23, Issue 12
  • DOI: 10.1063/1.3671736