基于人工神经网络的可压缩湍流大涡模拟模型

谢晨月; 王建春; 万敏平; 陈十一

doi:10.7527/S1000-6893.2021.25723

航空学报 >

2021 , Vol. 42 >Issue 9: 625723 - 625723

DOI: https://doi.org/10.7527/S1000-6893.2021.25723

国家数值风洞(NNW)进展及应用专栏

基于人工神经网络的可压缩湍流大涡模拟模型

谢晨月 ,
王建春 ,
万敏平 ,
陈十一

展开

1. 南方科技大学工学院力学与航空航天工程系, 深圳 518055;
2. 南方科技大学粤港澳数据驱动下的流体力学与工程应用联合实验室, 深圳 518055

收稿日期: 2021-03-30

修回日期: 2021-04-27

网络出版日期: 2021-05-20

基金资助

国家数值风洞工程；国家自然科学基金（91952104，92052301，91752201）；广东省科学技术厅项目（2020B1212030001）；深圳市科技创新委员会项目（KQTD20180411143441009）；南方科技大学科学与工程计算中心项目

收起

Artificial neural network model for large-eddy simulation of compressible turbulence

XIE Chenyue ,
WANG Jianchun ,
WAN Minping ,
CHEN Shiyi

Expand

1. Department of Mechanics and Aerospace Engineering, College of Engineering, Southern University of Science and Technology, Shenzhen 518055, China;
2. Guangdong-Hong Kong-Macao Joint Laboratory for Data-Driven Fluid Mechanics and Engineering Applications, Southern University of Science and Technology, Shenzhen 518055, China

Received date: 2021-03-30

Revised date: 2021-04-27

Online published: 2021-05-20

Supported by

National Numerical Windtunnel Project; National Natural Science Foundation of China (91952104, 92052301, 91752201); Project of Department of Science and Technology of Guangdong Province (2020B1212030001); Project of Technology and Innovation Commission of Shenzhen Municipality (KQTD20180411143441009); Project of Center for Computational Science and Engineering of Southern University of Science and Technology

Fold

摘要

在国家数值风洞（NNW）工程项目的指导下，空间人工神经网络（SANN）模型被用于强可压缩湍流大涡模拟（LES）研究，其中流场的湍流马赫数分别为0.6、0.8、1.0。基于湍流的多尺度空间结构特性和人工神经网络方法发展的高精度空间神经网络（SANN）模型适用于不可压缩湍流和弱可压缩湍流。对于强可压缩湍流，流场中会出现激波结构，给大涡模拟带来了挑战。本文的研究结果表明：SANN模型适用于强可压缩湍流的大涡模拟。在先验分析中，SANN模型预测的亚格子应力和亚格子热流的相关系数超过0.995，远远高于梯度模型和近似反卷积模型等传统模型；传统模型的相对误差大于30%，而SANN模型在这方面有很大的改进，相对误差低于11%。在后验分析中，与隐式大涡模拟（ILES）、动态Smagorinsky模型（DSM）、动态混合模型（DMM）相比，SANN模型能更精确地预测能谱、各类湍流统计特性以及瞬态流动结构。因此，基于湍流多尺度空间结构特性的人工神经网络模型加深了人们对强可压缩湍流亚格子建模的认识，同时可以服务于NNW工程的流体力学模型构造。

关键词： 可压缩湍流; 大涡模拟; 人工神经网络; 湍流模型; 国家数值风洞(NNW)工程

本文引用格式

谢晨月 , 王建春 , 万敏平 , 陈十一 . 基于人工神经网络的可压缩湍流大涡模拟模型[J]. 航空学报, 2021 , 42(9) : 625723 -625723 . DOI: 10.7527/S1000-6893.2021.25723

Abstract

The Spatial Artificial Neural Network (SANN) model is applied to perform Large Eddy Simulations (LES) of highly compressible turbulence at high turbulent Mach numbers of 0.6, 0.8 and 1.0 under the National Numerical Windtunnel (NNW) Project. In our previous studies, we developed the SANN model for incompressible and weakly compressible turbulence based on multi-scale spatial structures of turbulence. However, generations of shock waves in highly compressible turbulence pose great challenges to LES. This paper discusses the applicability of the SANN models for LES of highly compressible turbulence. It has been demonstrated that the correlation coefficients of the SANN model can be larger than 0.995. The relative errors of the SANN model can be smaller than 11%, which are much smaller than those of the traditional gradient model and the approximate deconvolution model in an a priori analysis for highly compressible turbulence. In an a posteriori analysis, we make a comparison of the results of the SANN model, the Implicit Large Eddy Simulation (ILES), the Dynamic Smagorinsky Model (DSM) and the Dynamic Mixed Model (DMM). It is shown that the SANN model performs better in the prediction of spectra and statistical properties of velocity and temperature, and instantaneous flow structures for highly compressible turbulence. The artificial neural network model with consideration of spatial features can deepen our understanding of subgrid-scale modeling for LES of highly compressible turbulence. At the same time, the model can contribute to the construction of the turbulence models of the NNW Project.

Key words： compressible turbulence; large eddy simulation; artificial neural network; turbulence model; National Numerical Windtunnel (NNW) Project

参考文献

[1] MOIN P, SQUIRES K, CABOT W, et al. A dynamic subgrid-scale model for compressible turbulence and scalar transport[J]. Physics of Fluids A:Fluid Dynamics, 1991, 3(11):2746-2757.
[2] MENEVEAU C, KATZ J. Scale-invariance and turbulence models for large-eddy simulation[J]. Annual Review of Fluid Mechanics, 2000, 32(1):1-32.
[3] CHEN S Y, XIA Z H, PEI S Y, et al. Reynolds-stress-constrained large-eddy simulation of wall-bounded turbulent flows[J]. Journal of Fluid Mechanics, 2012, 703:1-28.
[4] 张兆顺, 崔桂香, 许春晓. 湍流大涡数值模拟的理论和应用[M]. 北京:清华大学出版社, 2008. ZHANG Z S, CUI G X, XU C X. Theory and applica tion of large eddy simulation of turbulent flow[M]. Beijing:Tsinghua University Press, 2008(in Chinese).
[5] 傅德薰, 马延文, 李新亮. 可压缩湍流直接数值模拟[M]. 北京:科学出版社, 2010. FU D X, MA Y W, LI X L. Direct numerical simulation of compressible turbulence[M]. Beijing:Science Press, 2010(in Chinese).
[6] GARNIER E, ADAMS N, SAGAUT P. Large eddy simulation for compressible flows[M]. Dordrecht:Springer Netherlands, 2009.
[7] POPE S. Turbulent flows[M]. Cambridge:Cambridge University Press, 2000.
[8] SMAGORINSKY J. General circulation experiments with the primitive equations[J]. Monthly Weather Review, 1963, 91(3):99-164.
[9] LILLY D K. The representation of small-scale turbulence in numerical simulation experiments[C]//Proceedings of the IBM Scientific Computing Symposium on Environmental Sciences, 1967:195-210.
[10] DEARDORFF J W. A numerical study of three-dimensional turbulent channel flow at large Reynolds numbers[J]. Journal of Fluid Mechanics, 1970, 41(2):453-480.
[11] WANG J C, SHI Y P, WANG L P, et al. Effect of shocklets on the velocity gradients in highly compressible isotropic turbulence[J]. Physics of Fluids, 2011, 23(12):125103.
[12] WANG J C, SHI Y P, WANG L P, et al. Effect of com pressibility on the small-scale structures in isotropic turbulence[J]. Journal of Fluid Mechanics, 2012, 713:588-631.
[13] WANG J C, GOTOH T, WATANABE T. Spectra andstatistics in compressible isotropic turbulence[J]. Physi cal Review Fluids, 2017, 2:013403.
[14] WANG J C, GOTOH T, WATANABE T. Shocklet statistics in compressible isotropic turbulence[J]. Physical Review Fluids, 2017, 2(2):023401.
[15] WANG J C, GOTOH T, WATANABE T. Scaling and intermittency in compressible isotropic turbulence[J]. Physical Review Fluids, 2017, 2(5):053401.
[16] WANG J C, WAN M P, CHEN S, et al. Kinetic energy transfer in compressible isotropic turbulence[J]. Journal of Fluid Mechanics, 2018, 841:581-613.
[17] WANG J C, WAN M P, CHEN S, et al. Cascades of temperature and entropy fluctuations in compressible turbulence[J]. Journal of Fluid Mechanics, 2019, 867:195-215.
[18] 陈坚强. 国家数值风洞(NNW)工程关键技术研究进展[J/OL]. 中国科学:技术科学,(2021-04-28)[2021-05-10]. https://kns.cnki.net/kcms/detail/11.5844.TH.20210-428.0914.006.html. CHEN J Q. Advances in the key technologies of Chinese national numerical windtunnel project.Scientia Sinica Technologica, (2021-04-28)[2021-05-10]. https://kns.cnki.net/kcms/detail/11.5844.TH.20210428.0914.006.html (in Chinese).
[19] HUGHES T J R, MAZZEI L, JANSEN K E. Large Eddy Simulation and the variational multiscale method[J]. Computing and Visualization in Science, 2000, 3(1-2):47-59.
[20] EYINK G L. Locality of turbulent cascades[J]. Physica D:Nonlinear Phenomena, 2005, 207(1-2):91-116.
[21] LING J L, KURZAWSKI A, TEMPLETON J. Reynolds averaged turbulence modelling using deep neural networks with embedded invariance[J]. Journal of Fluid Mechanics, 2016, 807:155-166.
[22] XIAO H, WU J L, WANG J X, et al. Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier-Stokes simulations:A data-driven, physics-informed Bayesian approach[J]. Journal of Computational Physics, 2016, 324:115-136.
[23] DURAISAMY K, IACCARINO G, XIAO H. Turbulence modeling in the age of data[J]. Annual Review of Fluid Mechanics, 2019, 51(1):357-377.
[24] SARGHINI F, DE FELICE G, SANTINI S. Neural networks based subgrid scale modeling in large eddy simulations[J]. Computers & Fluids, 2003, 32(1):97-108.
[25] GAMAHARA M, HATTORI Y. Searching for turbulence models by artificial neural network[J]. Physical Review Fluids, 2017, 2(5):054604.
[26] VOLLANT A, BALARAC G, CORRE C. Subgrid-scale scalar flux modelling based on optimal estimation theory and machine-learning procedures[J]. Journal of Turbulence, 2017, 18(9):854-878.
[27] KUTZ J N. Deep learning in fluid dynamics[J]. Journal of Fluid Mechanics, 2017, 814:1-4.
[28] MAULIK R, SAN O. A neural network approach for the blind deconvolution of turbulent flows[J]. Journal of Fluid Mechanics, 2017, 831:151-181.
[29] JOFRE L, DOMINO S P, IACCARINO G. A framework for characterizing structural uncertainty in large-eddy simulation closures[J]. Flow, Turbulence and Combustion, 2018, 100(2):341-363.
[30] ZHOU Z D, WANG S Z, JIN G D. A structural subgrid-scale model for relative dispersion in large-eddy simulation of isotropic turbulent flows by coupling kinematic simulation with approximate deconvolution method[J]. Physics of Fluids, 2018, 30(10):105110.
[31] WANG Z, LUO K, LI D, et al. Investigations of data-driven closure for subgrid-scale stress in large-eddy simulation[J]. Physics of Fluids, 2018, 30(12):125101.
[32] MAULIK R, SAN O, RASHEED A, et al. Data-driven deconvolution for large eddy simulations of Kraichnan turbulence[J]. Physics of Fluids, 2018, 30(12):125109.
[33] ZHOU Z D, HE G W, WANG S Z, et al. Subgrid-scale model for large-eddy simulation of isotropic turbulent flows using an artificial neural network[J]. Computers & Fluids, 2019, 195:104319.
[34] ZHU L Y, ZHANG W W, KOU J Q, et al. Machine learning methods for turbulence modeling in subsonic flows around airfoils[J]. Physics of Fluids, 2019, 31(1):015105.
[35] MAULIK R, SAN O, RASHEED A, et al. Subgrid modelling for two-dimensional turbulence using neural networks[J]. Journal of Fluid Mechanics, 2019, 858:122-144.
[36] YU M, HUANG W X, XU C X, Data-driven construction of a reduced-order model for supersonic boundary layer transition[J]. Journal of Fluid Mechanics, 2019, 874:1096-1114.
[37] MA C, WANG J C, E W N. Model reduction with memory and the machine learning of dynamical systems[J]. Communications in Computational Physics, 2019, 25(4):947-962.
[38] XIE C Y, WANG J C, LI K, et al. Artificial neural network approach to large-eddy simulation of compressible isotropic turbulence[J]. Physical Review E, 2019, 99(5-1):053113.
[39] XIE C Y, LI K, MA C, et al. Modeling subgrid-scale force and divergence of heat flux of compressible isotropic turbulence by artificial neural network[J]. Physical Review Fluids, 2019, 4(10):104605.
[40] XIE C Y, WANG J C, LI H, et al. Artificial neural network mixed model for large eddy simulation of compressible isotropic turbulence[J]. Physics of Fluids, 2019, 31(8):085112.
[41] XIE C Y, WANG J C, LI H, et al. Spatial artificial neural network model for subgrid-scale stress and heat flux of compressible turbulence[J]. Theoretical and Applied Mechanics Letters, 2020, 10(1):27-32.
[42] XIE C Y, WANG J C, LI H, et al. Spatially multi-scale artificial neural network model for large eddy simulation of compressible isotropic turbulence[J]. AIP Advances, 2020, 10(1):015044.
[43] XIE C Y, WANG J C, E W N. Modeling subgrid-scale forces by spatial artificial neural networks in large eddy simulation of turbulence[J]. Physical Review Fluids, 2020, 5:054606.
[44] XIE C Y, YUAN Z L, WANG J C. Artificial neural network-based nonlinear algebraic models for large eddy simulation of turbulence[J]. Physics of Fluids, 2020, 32(11):115101.
[45] YUAN Z L, XIE C Y, WANG J C. Deconvolutional artificial neural network models for large eddy simulation of turbulence[J]. Physics of Fluids, 2020, 32(11):115106.
[46] 谢晨月, 袁泽龙, 王建春, 等. 基于人工神经网络的湍流大涡模拟方法[J]. 力学学报, 2021, 53(1):1-16. XIE C Y, YUAN Z L, WANG J C, et al. Artificial neural network-based subgrid-scale models for large-eddy simulation of turbulence[J]. Chinese Journal of Theoretical and Applied Mechanics, 2021, 53(1):1-16(in Chinese).
[47] XIE C Y, XIONG X M, WANG J C. Artificial neural network approach for turbulence models:A local framework[DB/OL]. arXiv preprint:2101.10528, 2021.
[48] 任峰, 高传强, 唐辉. 机器学习在流动控制领域的应用及发展趋势[J]. 航空学报, 2021, 42(4):524686. REN F, GAO C Q, TANG H. Machine learning for flow control:Applications and development trends[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4):524686(in Chinese).
[49] 张伟伟, 寇家庆, 刘溢浪. 智能赋能流体力学展望[J]. 航空学报, 2021, 42(4):524689. ZHANG W W, KOU J Q, LIU Y L. Prospect of artifi-cial intelligence empowered fluid mechanics[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4):524689(in Chinese).
[50] 何创新, 邓志文, 刘应征. 湍流数据同化技术及应用[J]. 航空学报, 2021, 42(4):524704. HE C X, DENG Z W, LIU Y Z. Turbulent flow data assimilation and its applications[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4):524704(in Chinese).
[51] 叶舒然, 张珍, 王一伟, 等. 基于卷积神经网络的深度学习流场特征识别及应用进展[J]. 航空学报, 2021, 42(4):524736. YE S R, ZHANG Z, WANG Y W, et al. Progress in deep convolutional neural network based flow field recognition and its applications[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4):524736(in Chinese).
[52] 李霓, 布树辉, 尚柏林, 等. 飞行器智能设计愿景与关键问题[J]. 航空学报, 2021, 42(4):524752. LI N, BU S H, SHANG B L, et al. Aircraft intelligent design:Visions and key technologies[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(4):524752(in Chinese).
[53] SAMTANEY R, PULLIN D I, KOSOVIĆ B. Direct numerical simulation of decaying compressible turbulence and shocklet statistics[J]. Physics of Fluids, 2001, 13(5):1415-1430.
[54] FAVRE A. Equations des gaz turbulents compressible.I. Formes generals[J]. Journal de Mcanique, 1965, 4:361-390.
[55] VREMAN B, GEURTS B, KUERTEN H. Subgrid-modelling in LES of compressible flow[J]. Applied Scientific Research, 1995, 54(3):191-203.
[56] MARTÍN P M, PIOMELLI U, CANDLER G V. Subgrid-scale models for compressible large-eddy simulations[J]. Theoretical and Computational Fluid Dynamics, 2000, 13(5):361-376.
[57] XIE C Y, WANG J C, LI H, et al. A modified optimal LES model for highly compressible isotropic turbulence[J]. Physics of Fluids, 2018, 30(6):065108.
[58] CHAI X C, MAHESH K. Dynamic-equation model for large-eddy simulation of compressible flows[J]. Journal of Fluid Mechanics, 2012, 699:385-413.
[59] WANG J, WANG L P, XIAO Z, et al. A hybrid numerical simulation of isotropic compressible turbulence[J]. Journal of Computational Physics, 2010, 229(13):5257-5279.
[60] LELE S K. Compact finite difference schemes with spectral-like resolution[J]. Journal of Computational Physics, 1992, 103(1):16-42.
[61] BALSARA D S, SHU C W. Monotonicity preserving weighted essentially non-oscillatory schemes with increasingly high order of accuracy[J]. Journal of Computational Physics, 2000, 160(2):405-452.
[62] GHOSAL S. An analysis of numerical errors in large-eddy simulations of turbulence[J]. Journal of Computational Physics, 1996, 125(1):187-206.
[63] MEYERS J, GEURTS B J, BAELMANS M. Database analysis of errors in large-eddy simulation[J]. Physics of Fluids, 2003, 15(9):2740-2755.
[64] CHOW F K, MOIN P. A further study of numerical errors in large-eddy simulations[J]. Journal of Computational Physics, 2003, 184(2):366-380.
[65] ALUIE H, EYINK G L. Localness of energy cascade in hydrodynamic turbulence. II. Sharp spectral filter[J]. Physics of Fluids, 2009, 21(11):115108.
[66] ALUIE H. Compressible turbulence:The cascade and its locality[J]. Physical Review Letters, 2011, 106(17):174502.
[67] KINGMA D, BA J. Adam:A method for stochastic optimization[DB/OL]. arXiv preprint:1412.6980, 2014.
[68] BARDINA J, FERZIGER J, REYNOLDS W. Improved subgrid-scale models for large-eddy simulation:AIAA-1980-1357[R]. Reston:AIAA, 1980.
[69] LIU S W, MENEVEAU C, KATZ J. On the properties of similarity subgrid-scale models as deduced from measurements in a turbulent jet[J]. Journal of Fluid Mechanics, 1994, 275:83-119.
[70] STOLZ S, ADAMS N A. An approximate deconvolution procedure for large-eddy simulation[J]. Physics of Fluids, 1999, 11(7):1699-1701.
[71] STOLZ S, ADAMS N A, KLEISER L. The approximate deconvolution model for large-eddy simulations of compressible flows and its application to shock-turbulent-boundary-layer interaction[J]. Physics of Fluids, 2001, 13(10):2985-3001.
[72] LILLY D K. A proposed modification of the Germano subgrid-scale closure method[J]. Physics of Fluids A:Fluid Dynamics, 1992, 4(3):633-635.
[73] GERMANO M. Turbulence:The filtering approach[J]. Journal of Fluid Mechanics, 1992, 238:325-336.
[74] SHI Y P, XIAO Z L, CHEN S Y. Constrained subgrid-scale stress model for large eddy simulation[J]. Physics of Fluids, 2008, 20(1):011701.
[75] MENEVEAU C. Statistics of turbulence subgrid-scale stresses:Necessary conditions and experimental tests[J]. Physics of Fluids, 1994, 6(2):815-833.
[76] STEVENS R J A M, WILCZEK M, MENEVEAU C. Large-eddy simulation study of the logarithmic law for second-and higher-order moments in turbulent wall-bounded flow[J]. Journal of Fluid Mechanics, 2014, 757:880-907.
[77] BUZZICOTTI M, LINKMANN M, ALUIE H, et al. Effect of filter type on the statistics of energy transfer between resolved and subfilter scales from a-priori analysis of direct numerical simulations of isotropic turbulence[J]. Journal of Turbulence, 2018, 19(2):167-197.
[78] LINKMANN M, BUZZICOTTI M, BIFERALE L. Multi-scale properties of large eddy simulations:Corre lations between resolved-scale velocity-field increments and subgrid-scale quantities[J]. Journal of Turbulence, 2018, 19(6):493-527.

Options

文章导航

模态框（Modal）标题

摘要

本文引用格式

Abstract

参考文献