×

zbMATH — the first resource for mathematics

Using machine learning to detect the turbulent region in flow past a circular cylinder. (English) Zbl 07271145
Summary: Detecting the turbulent/non-turbulent interface is a challenging topic in turbulence research. In the present study, machine learning methods are used to train detectors for identifying turbulent regions in the flow past a circular cylinder. To ensure that the turbulent/non-turbulent interface is independent of the reference frame of coordinates and is physics-informed, we propose to use invariants of tensors appearing in the transport equations of velocity fluctuations, strain-rate tensor and vortical tensor as the input features to identify the flow state. The training samples are chosen from numerical simulation data at two Reynolds numbers, \(Re=100\) and 3900. Extreme gradient boosting (XGBoost) is utilized to train the detector, and after training, the detector is applied to identify the flow state at each point of the flow field. The trained detector is found robust in various tests, including the applications to the entire fields at successive snapshots and at a higher Reynolds number \(Re=5000\). The objectivity of the detector is verified by changing the input features and the flow region for choosing the turbulent training samples. Compared with the conventional methods, the proposed method based on machine learning shows its novelty in two aspects. First, no threshold value needs to be specified explicitly by the users. Second, machine learning can treat multiple input variables, which reflect different properties of turbulent flows, including the unsteadiness, vortex stretching and three-dimensionality. Owing to these advantages, XGBoost generates a detector that is more robust than those obtained from conventional methods.
MSC:
76 Fluid mechanics
Software:
XGBoost
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] Alsalman, M., Colvert, B. & Kanso, E.2019Training bioinspired sensors to classify flows. Bioinspir. Biomim.14, 016009.
[2] Anand, R. K., Boersma, B. J. & Agrawal, A.2009Detection of turbulent/non-turbulent interface for an axisymmetric turbulent jet: evaluation of known criteria and proposal of a new criterion. Exp. Fluids47, 995-1007.
[3] Bisset, D. K., Hunt, J. C. R. & Rogers, M. M.2002The turbulent/non-turbulent interface bounding a far wake. J. Fluid Mech.451, 383-410. · Zbl 1156.76397
[4] Borrell, G. & Jimémez, J.2016Properties of the turbulent/non-turbulent interface in boundary layers. J. Fluid Mech.451, 383-410.
[5] Chauhan, K., Philip, J., De Silva, C. M., Hutchins, N. & Marusic, I.2014The turbulent/non-turbulent interface and entrainment in a boundary layer. J. Fluid Mech.742, 119-151.
[6] Chen, T. & Guestrin, C.2016 XGBoost: a scalable tree boosting system. In ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM Press.
[7] Colvert, B., Alsalman, M. & Kanso, E.2018Classifying vortex wakes using neural networks. Bioinspir. Biomim.13, 025003.
[8] Corrsin, S. & Kistler, A. L.1954 Free-stream boundaries of turbulent flows. Technical Report Archive & Image Library.
[9] Cui, Z., Yang, Z., Jiang, H.-Z., Huang, W.-X. & Shen, L.2018A sharp-interface immersed boundary method for simulating incompressible flows with arbitrarily deforming smooth boundaries. Intl J. Comput. Methods15, 1750080. · Zbl 1404.76160
[10] Duraisamy, K., Iaccarino, G. & Xiao, H.2019Turbulence modeling in the age of data. Annu. Rev. Fluid Mech.51, 357-377. · Zbl 1412.76040
[11] Fukami, K., Fukagata, K. & Taira, K.2019Super-resolution reconstruction of turbulent flows with machine learning. J. Fluid Mech.870, 106-120.
[12] Gamahara, M. & Hattori, Y.2017Searching for turbulence models by artificial neural network. Phys. Rev. Fluids2, 054604.
[13] Germano, M., Piomelli, U., Moin, P. & Cabot, W. H.1991A dynamic subgrid-scale eddy viscosity model. Phys. Fluids A3 (7), 1760-1765. · Zbl 0825.76334
[14] Green, M. A., Rowley, C. W. & Haller, G.2007Detection of Lagragian coherent strucutres in 3D turbulence. J. Fluid Mech.572, 111-120. · Zbl 1111.76025
[15] Haller, G.2002Lagrangian coherent strucutres from approximate velocity data. Phys. Fluids14, 1851-1861.
[16] Huang, J., Liu, H. & Cai, W.2019Online in situ prediction of 3-D flame evolution from its history 2-D projections via deep learning. J. Fluid Mech.875, R2. · Zbl 1419.76354
[17] Hunt, J. C. R., Wray, A. A. & Moin, P.1988 Eddies, streams, and convergene zones in turbulent flows. In Proceeding of the Summer Program, Center for Turbulence Research, pp. 193-208. Stanford University/NASA.
[18] Jeong, J. & Hussain, F.1995On the identification of a vortex. J. Fluid Mech.285, 69-94. · Zbl 0847.76007
[19] Johansson, P. B. V. & George, W. K.2003Equilibrium similarity, effects of initial conditions and local Reynolds number on the axisymmetric wake. Phys. Fluids15, 603-617. · Zbl 1185.76187
[20] Kravchenko, A. G. & Moin, P.2000Numerical studies of flow over a circular cylinder at \(Re_D=3900\). Phys. Fluids12, 403-417. · Zbl 1149.76441
[21] Lee, J. & Zaki, T. A.2018Detection algorithm for turbulent interfaces and large-scale structures in intermittent flows. Comput. Fluids175, 142-158. · Zbl 1410.76242
[22] Lee, S. & You, D.2019Data-driven prediction of unsteady flow over a circular cylinder using deep learning. J. Fluid Mech.879, 217-254. · Zbl 1430.76311
[23] Lilly, D. K.1992A proposed modification of the Germano subgrid scale closure method. Phys. Fluids A4 (3), 633-635.
[24] Ling, J., Jones, R. & Templeton, J.2016aMachine learning strategies for systems with invariance properties. J. Comput. Phys.318, 22-35. · Zbl 1349.76124
[25] Ling, J., Kurzawski, A. & Templeton, J.2016bReynolds averaged turbulence modelling using deep neural networks with embedded invariance. J. Fluid Mech.807, 155-166. · Zbl 1383.76175
[26] Ling, J. & Templeton, J.2015Evaluation of machine learning algorithms for prediction of regions of high Reynolds averaged Navier Stokes uncertainty. Phys. Fluids27, 085103.
[27] Ma, M., Lu, J. & Tryggvason, G.2015Using statistical learning to close two-fluid multiphase flow equations for a simple bubbly system. Phys. Fluids27, 092101.
[28] Ma, X., Karamanos, G.-S. & Karniadakis, G. E.2000Dynamics and low-dimensionality of a turbulent near wake. J. Fluid Mech.410, 29-65. · Zbl 0987.76041
[29] Maulik, R. & San, O.2017A neural network approach for the blind deconvolution of turbulent flows. J. Fluid Mech.831, 151-181. · Zbl 1421.76134
[30] Maulik, R., San, O., Rasheed, A. & Vedula, P.2018Data-driven deconvolution for large eddy simulations of Kraichnan turbulence. Phys. Fluids30, 125109.
[31] Nolan, K. P. & Zaki, T. A.2013Conditional sampling of transitional boundary layers in pressure gradients. J. Fluid Mech.728, 306-339. · Zbl 1291.76106
[32] Parish, E. J. & Duraisamy, K.2016A paradigm for data-driven predictive modeling using field inversion and machine learning. J. Comput. Phys.305, 758-774. · Zbl 1349.76006
[33] Rehill, B., Walsh, E. J., Brandt, L., Schlatter, P. & Zaki, T. A.2013Identifying turbulent spots in transitional boundary layers. Trans. ASME: J. Turbomach.135, 011019.
[34] De Silva, C. M., Philip, J., Chauhan, K., Meneveau, C. & Marusic, I.2013Multiscale geometry and scaling of the turbulent-nonturbulent interface in high Reynolds number boundary layers. Phys. Rev. Lett.111, 044501.
[35] Ströfer, C. M., Wu, J.-L., Xiao, H. & Paterson, E.2019Data-driven, physics-based feature extraction from fluid flow fields using convolutional neural networks. Commun. Comput. Phys.25, 625-650.
[36] Vollant, A., Balarac, G. & Corre, C.2017Subgrid-scale scalar flux modelling based on optimal estimation theory and machine-learning procedures. J. Turbul.18, 854-878.
[37] Wang, J.-X., Wu, J.-L. & Xiao, H.2017Physics-informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data. Phys. Rev. Fluids2, 034603.
[38] Wang, Z., Luo, K., Li, D., Tan, J. H. & Fan, J. R.2018Investigations of data-driven closure for subgrid-scale stress in large-eddy simulation. Phys. Fluids30, 125101.
[39] Westerweel, J., Fukushima, C., Pedersen, J. M. & Hunt, J. C. R.2009Momentum and scalar transport at the turbulent/non-turbulent interface of a jet. J. Fluid Mech.631, 199-230. · Zbl 1181.76015
[40] Williamson, C. H. K.1996Vortex dynamics in the cylinder wake. Annu. Rev. Fluid Mech.28, 477-539.
[41] Wu, J., Xiao, H., Sun, R. & Wang, Q.2019aReynolds-averaged Navier-Stokes equations with explicit data-driven Reynolds stress closure can be ill-conditioned. J. Fluid Mech.869, 553-586. · Zbl 1429.76066
[42] Wu, J.-L., Xiao, H. & Paterson, E.2018Physics-informed machine learning approach for augmenting turbulence models: a comprehensive framework. Phys. Rev. Fluids3, 074602.
[43] Wu, Z., Lee, J., Meneveau, C. & Zaki, T.2019bApplication of a self-organizing map to identify the turbulent-boundary-layer interface in a transitional flow. Phys. Rev. Fluids4, 023902.
[44] Xiao, H., Wu, J.-L., Wang, J.-X., Sun, R. & Roy, C. J.2016Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier-Stokes simulations: a data-driven physics-informed Bayesian approach. J. Comput. Phys.324, 115-136. · Zbl 1371.76082
[45] Zhou, Z., He, G., Wang, S. & Jin, G.2019Subgird-scale model for large-eddy simulation of isotropic turbulent flows using an artificial neural network. Comput. Fluids195, 104319. · Zbl 07127685
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.