Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
156 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Investigation of Deep Learning-Based Filtered Density Function for Large Eddy Simulation of Turbulent Scalar Mixing (2307.07453v2)

Published 14 Jul 2023 in physics.flu-dyn and cs.CE

Abstract: A filtered density function (FDF) model based on deep neural network (DNN), termed DNN-FDF, is introduced for large eddy simulation (LES) of turbulent flows involving conserved scalar transport. The primary objectives of this study are to develop the DNN-FDF models and evaluate their predictive capability in accounting for various filtered moments, including that of non-linear source terms. A systematic approach is proposed to select DNN training sample size and architecture via learning curves to minimize bias and variance. Two DNN-FDF models are developed, one utilizing FDF data from Direct Numerical Simulations (DNS) of constant-density temporal mixing layer, and the other from zero-dimensional pairwise mixing stirred reactor simulations. The latter is particularly intended for cases where generating DNS data is computationally infeasible. DNN-FDF models are applied for LES of a variable-density temporal mixing layer. The accuracy and consistency of both DNN-FDF models are established by comparing their predicted filtered scalar moments with those of conventional LES, where moment transport equations are directly solved. The DNN-FDF models are shown to outperform a widely used presumed-FDF model, especially for multi-modal FDFs and higher variance values. Results are further assessed against DNS and the transported FDF method. The latter couples LES with Monte Carlo for mixture fraction FDF computation. Most importantly, the study shows that DNN-FDF models can accurately filter highly non-linear functions within variable-density flows, highlighting their potential for turbulent reacting flow simulations. Overall, the DNN-FDF approach is shown to offer an accurate yet computationally economical approach for describing turbulent scalar transport.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (47)
  1. P. Givi, “Filtered density function for subgrid scale modeling of turbulent combustion,” AIAA Journal 44, 16–23 (2006), https://doi.org/10.2514/1.15514 .
  2. S. B. Pope, “A Monte Carlo method for the PDF equations of turbulent reactive flow,” Combustion Science and Technology 25, 159–174 (1981).
  3. S. B. Pope, Turbulent flows (Cambridge University Press, Cambridge, UK, 2000).
  4. S. B. Pope, “PDF methods for turbulent reactive flows,” Progress in Energy and Combustion Science 11, 119–192 (1985).
  5. S. S. Girimaji, “Assumed β𝛽\betaitalic_β-PDF model for turbulent mixing: Validation and extension to multiple scalar mixing,” Combustion Science and Technology 78, 177–196 (1991).
  6. C. Madnia and P. Givi, ‘‘Direct numerical simulation and large eddy simulation of reacting homogeneous turbulence,” Large Eddy Simulations of Complex Engineering and Geophysical Flows 5, 315–346 (1993).
  7. J. Jiménez, A. Liñán, M. M. Rogers,  and F. J. Higuera, “A priori testing of subgrid models for chemically reacting non-premixed turbulent shear flows,” Journal of Fluid Mechanics 349, 149–171 (1997).
  8. A. Klimenko and R. Bilger, “Conditional moment closure for turbulent combustion,” Progress in Energy and Combustion Science 25, 595–687 (1999).
  9. N. Peters, “Laminar diffusion flamelet models in non-premixed turbulent combustion,” Progress in Energy and Combustion Science 10, 319–339 (1984).
  10. J. v. Oijen and L. D. Goey, “Modelling of premixed laminar flames using flamelet-generated manifolds,” Combustion Science and Technology 161, 113–137 (2000).
  11. C. D. Pierce and P. Moin, “Progress-variable approach for large-eddy simulation of non-premixed turbulent combustion,” Journal of Fluid Mechanics 504, 73–97 (2004).
  12. P. Wang, N. Platova, J. Fröhlich,  and U. Maas, “Large eddy simulation of the preccinsta burner,” International Journal of Heat and Mass Transfer 70, 486–495 (2014).
  13. J. Floyd, A. M. Kempf, A. Kronenburg,  and R. H. Ram, “A simple model for the filtered density function for passive scalar combustion LES,” Combustion Theory and Modelling 13, 559–588 (2009).
  14. A. W. Cook and J. J. Riley, “A subgrid model for equilibrium chemistry in turbulent flows,” Physics of Fluids 6, 2868–2870 (1994).
  15. H. Wang, P. Zhang,  and J. Tao, “Examination of probability distribution of mixture fraction in LES/FDF modelling of a turbulent partially premixed jet flame,” Combustion Theory and Modelling  (2021), 10.1080/13647830.2021.2016980.
  16. S. Yun, M. F. Lightstone,  and M. J. Thomson, “An evaluation of β𝛽\betaitalic_β PDF integration using the density-weighted PDF and the un-weighted PDF,” International Journal of Thermal Sciences 44, 421–428 (2005).
  17. C. Tong, “Measurements of conserved scalar filtered density function in a turbulent jet,” Physics of Fluids 13, 2923–2937 (2001), https://doi.org/10.1063/1.1402171 .
  18. Z. Wang, K. Luo, D. Li, J. Tan,  and J. Fan, “Investigations of data-driven closure for subgrid-scale stress in large-eddy simulation,” Physics of Fluids 30, 125101 (2018), https://doi.org/10.1063/1.5054835 .
  19. A. K. Chatzopoulos and S. Rigopoulos, “A chemistry tabulation approach via rate-controlled constrained equilibrium (RCCE) and artificial neural networks (ANNs), with application to turbulent non-premixed CH44{}_{4}start_FLOATSUBSCRIPT 4 end_FLOATSUBSCRIPT/H22{}_{2}start_FLOATSUBSCRIPT 2 end_FLOATSUBSCRIPT/N22{}_{2}start_FLOATSUBSCRIPT 2 end_FLOATSUBSCRIPT flames,” Proceedings of the Combustion Institute 34, 1465–1473 (2013).
  20. S. Bansude, F. Imani,  and R. Sheikhi, “A data-driven framework for computationally efficient integration of chemical kinetics using neural ordinary differential equations,” ASME Open Journal of Engineering 2 (2023a), 10.1115/1.4062105, 021022.
  21. O. Owoyele and P. Pal, “ChemNODE: A neural ordinary differential equations framework for efficient chemical kinetic solvers,” Energy and AI , 100118 (2021).
  22. H.-T. Nguyen, P. Domingo, L. Vervisch,  and P.-D. Nguyen, “Machine learning for integrating combustion chemistry in numerical simulations,” Energy and AI 5, 100082 (2021).
  23. L. L. Franke, A. K. Chatzopoulos,  and S. Rigopoulos, “Tabulation of combustion chemistry via Artificial Neural Networks (ANNs): Methodology and application to LES-PDF simulation of Sydney flame L,” Combustion and Flame 185, 245–260 (2017).
  24. R. Ranade and T. Echekki, “A framework for data-based turbulent combustion closure: A posteriori validation,” Combustion and Flame 210, 279–291 (2019).
  25. S. Yao, B. Wang, A. Kronenburg,  and O. T. Stein, “Modeling of sub-grid conditional mixing statistics in turbulent sprays using machine learning methods,” Physics of Fluids 32 (2020), 10.1063/5.0027524.
  26. M. T. H. de Frahan, S. Yellapantula, R. King, M. S. Day,  and R. W. Grout, “Deep learning for presumed probability density function models,” Combustion and Flame 208, 436–450 (2019).
  27. K. M. Gitushi, R. Ranade,  and T. Echekki, “Investigation of deep learning methods for efficient high-fidelity simulations in turbulent combustion,” Combustion and Flame 236, 111814 (2022).
  28. F. A. Jaberi, P. J. Colucci, S. James, P. Givi,  and S. B. Pope, “Filtered mass density function for large-eddy simulation of turbulent reacting flows,” Journal of Fluid Mechanics 401, 85–121 (1999).
  29. M. R. H. Sheikhi, P. Givi,  and S. B. Pope, “Velocity-scalar filtered mass density function for large eddy simulation of turbulent reacting flows,” Physics of Fluids 19, 095106 (2007).
  30. P. J. Colucci, F. A. Jaberi, P. Givi,  and S. B. Pope, “Filtered density function for large eddy simulation of turbulent reacting flows,” Physics of Fluids 10, 499–515 (1998).
  31. T. G. Drozda, M. R. H. Sheikhi, C. K. Madnia,  and P. Givi, “Developments in formulation and application of the filtered density function,” Flow Turbul. Combust. 78, 35–67 (2007).
  32. H. Pitsch, “Large-Eddy Simulation of Turbulent Combustion,” Annu. Rev. Fluid Mech. 38, 453–482 (2006).
  33. D. C. Haworth, “Progress in probability density function methods for turbulent reacting flows,” Prog. Energ. Combust. 36, 168–259 (2010).
  34. Z. X. Chen, I. Langella,  and N. Swaminathan, “The role of cfd in modern jet engine combustor design,” in Environmental Impact of Aviation and Sustainable Solutions, edited by R. K. Agarwal (IntechOpen, Rijeka, 2019) Chap. 3.
  35. C. W. Gardiner, Handbook of Stochastic Methods (Springer-Verlag, New York, NY, 1990).
  36. E. E. O’Brien, “The probability density function (PDF) approach to reacting turbulent flows,” in Turbulent Reacting Flows, Topics in Applied Physics, Vol. 44, edited by P. A. Libby and F. A. Williams (Springer-Verlag, Heidelberg, 1980) Chap. 5, pp. 185–218.
  37. C. Dopazo and E. E. O’Brien, “Statistical treatment of non-isothermal chemical reactions in turbulence,” Combust. Sci. and Tech. 13, 99–112 (1976).
  38. M. R. Sheikhi, T. G. Drozda, P. Givi,  and S. B. Pope, “Velocity-scalar filtered density function for large eddy simulation of turbulent flows,” Physics of Fluids 15, 2321–2337 (2003).
  39. B. Yang and S. Pope, “An investigation of the accuracy of manifold methods and splitting schemes in the computational implementation of combustion chemistry,” Combustion and Flame 112, 16–32 (1998).
  40. M. A. Singer, S. B. Pope,  and H. N. Najm, “Operator-splitting with isat to model reacting flow with detailed chemistry,” Combustion Theory and Modelling 10, 199–217 (2006), https://doi.org/10.1080/13647830500307501 .
  41. F. Hadi, M. Janbozorgi, M. R. H. Sheikhi,  and H. Metghalchi, “A study of interactions between mixing and chemical reaction using the rate-controlled constrained-equilibrium method,” Journal of Non-Equilibrium Thermodynamics 41, 257–278 (2016).
  42. S. Bansude, F. Imani,  and R. Sheikhi, “Performance assessment of chemical kinetics neural ordinary differential equations in pairwise mixing stirred reactor,” ASME Open Journal of Engineering 2 (2023b), 10.1115/1.4056476, 021008.
  43. J. Heaton, “Ian goodfellow, yoshua bengio, and aaron courville: Deep learning: The MIT press, 2016, 800 pp, isbn: 0262035618,” Genetic Programming and Evolvable Machines 19, 305–307 (2018).
  44. A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga, A. Desmaison, A. Kopf, E. Yang, Z. DeVito, M. Raison, A. Tejani, S. Chilamkurthy, B. Steiner, L. Fang, J. Bai,  and S. Chintala, “Pytorch: An imperative style, high-performance deep learning library,” in Advances in Neural Information Processing Systems 32 (Curran Associates, Inc., 2019) pp. 8024–8035.
  45. P. Virtanen, R. Gommers, T. E. Oliphant, M. Haberland, T. Reddy, D. Cournapeau, E. Burovski, P. Peterson, W. Weckesser, J. Bright, S. J. van der Walt, M. Brett, J. Wilson, K. J. Millman, N. Mayorov, A. R. J. Nelson, E. Jones, R. Kern, E. Larson, C. J. Carey, İ. Polat, Y. Feng, E. W. Moore, J. VanderPlas, D. Laxalde, J. Perktold, R. Cimrman, I. Henriksen, E. A. Quintero, C. R. Harris, A. M. Archibald, A. H. Ribeiro, F. Pedregosa, P. van Mulbregt,  and SciPy 1.0 Contributors, “SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python,” Nature Methods 17, 261–272 (2020).
  46. M. R. Sheikhi, T. G. Drozda, P. Givi, F. A. Jaberi,  and S. B. Pope, “Large eddy simulation of a turbulent nonpremixed piloted methane jet flame (Sandia Flame D),” Proceedings of the Combustion Institute 30, 549–556 (2005).
  47. G. L. Brown and A. Roshko, “On density effects and large structure in turbulent mixing layers,” Journal of Fluid Mechanics 64, 775–816 (1974).

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now