No CrossRef data available.
Article contents
Data-driven detached-eddy simulations based on explicit algebraic stress expressions for turbulent flows
Published online by Cambridge University Press: 12 January 2026
Abstract
This work proposes a data-driven explicit algebraic stress-based detached-eddy simulation (DES) method. Despite the widespread use of data-driven methods in model development for both Reynolds-averaged Navier–Stokes (RANS) and large-eddy simulations (LES), their applications to DES remain limited. The challenge mainly lies in the absence of modelled stress data, the requirement for proper length scales in RANS and LES branches, and the maintenance of a reasonable switching behaviour. The data-driven DES method is constructed based on the algebraic stress equation. The control of RANS/LES switching is achieved through the eddy viscosity in the linear part of the modelled stress, under the 
$\ell ^2-\omega$ DES framework. Three model coefficients associated with the pressure–strain terms and the LES length scale are represented by a neural network as functions of scalar invariants of velocity gradient. The neural network is trained using velocity data with the ensemble Kalman method, thereby circumventing the requirement for modelled stress data. Moreover, the baseline coefficient values are incorporated as additional reference data to ensure reasonable switching behaviour. The proposed approach is evaluated on two challenging turbulent flows, i.e. the secondary flow in a square duct and the separated flow over a bump. The trained model achieves significant improvements in predicting mean flow statistics compared with the baseline model. This is attributed to improved predictions of the modelled stress. The trained model also exhibits reasonable switching behaviour, enlarging the LES region to resolve more turbulent structures. Furthermore, the model shows satisfactory generalization capabilities for both cases in similar flow configurations.
JFM classification
Information
- Type
- JFM Papers
- Information
- Copyright
- © The Author(s), 2026. Published by Cambridge University Press
References
Abadi, M., et al. 2016 TensorFlow: Large-scale machine learning on heterogeneous systems. arXiv: 1603.04467.Google Scholar
Ashton, N., Revell, A., Prosser, R. & Uribe, J. 2013 Development of an alternative delayed detached-eddy simulation formulation based on elliptic relaxation. AIAA J. 51 (2), 513–519.10.2514/1.J051808CrossRefGoogle Scholar
Bader, S.H., Yin, Z. & Durbin, P.A. 2022 A hybrid model for turbulence and transition, with a locally varying coefficient. Flow Turbul. Combust. 108 (4), 935–954.10.1007/s10494-021-00308-4CrossRefGoogle Scholar
Balin, R. & Jansen, K.E. 2021 Direct numerical simulation of a turbulent boundary layer over a bump with strong pressure gradients. J. Fluid Mech. 918, A14.10.1017/jfm.2021.312CrossRefGoogle Scholar
Banerjee, S., Krahl, R., Durst, F. & Zenger, C. 2007 Presentation of anisotropy properties of turbulence, invariants versus eigenvalue approaches. J. Turbul. 8, N32.10.1080/14685240701506896CrossRefGoogle Scholar
Bose, R. & Roy, A.M. 2024 Invariance embedded physics-infused deep neural network-based sub-grid scale models for turbulent flows. Engng Applics Artif. Intell. 128, 107483.10.1016/j.engappai.2023.107483CrossRefGoogle Scholar
Brunton, S.L., Noack, B.R. & Koumoutsakos, P. 2020 Machine learning for fluid mechanics. Annu. Rev. Fluid Mech. 52 (1), 477–508.10.1146/annurev-fluid-010719-060214CrossRefGoogle Scholar
Chaouat, B. 2017 The state of the art of hybrid RANS/LES modeling for the simulation of turbulent flows. Flow Turbul. Combust. 99, 279–327.10.1007/s10494-017-9828-8CrossRefGoogle ScholarPubMed
Deck, S. 2012 Recent improvements in the zonal detached eddy simulation (ZDES) formulation. Theor. Comput. Fluid Dyn. 26, 523–550.10.1007/s00162-011-0240-zCrossRefGoogle Scholar
Duraisamy, K. 2021 Perspectives on machine learning-augmented Reynolds-averaged and large eddy simulation models of turbulence. Phys. Rev. Fluids 6 (5), 050504.10.1103/PhysRevFluids.6.050504CrossRefGoogle Scholar
Duraisamy, K., Iaccarino, G. & Xiao, H. 2019 Turbulence modeling in the age of data. Annu. Rev. Fluid Mech. 51, 357–377.10.1146/annurev-fluid-010518-040547CrossRefGoogle Scholar
Durbin, P.A. 2011 Statistical Theory and Modeling for Turbulent Flows. Wiley Sons Ltd.Google Scholar
Durbin, P.A. & Speziale, C.G. 1994 Realizability of second-moment closure via stochastic analysis. J. Fluid Mech. 280, 395–407.10.1017/S0022112094002983CrossRefGoogle Scholar
Durbin, P.A. 2018 Some recent developments in turbulence closure modeling. Annu. Rev. Fluid Mech. 50 (1), 77–103.10.1146/annurev-fluid-122316-045020CrossRefGoogle Scholar
Elnahhas, A. & Johnson, P.L. 2022 On the enhancement of boundary layer skin friction by turbulence: an angular momentum approach. J. Fluid Mech. 940, A36.10.1017/jfm.2022.264CrossRefGoogle Scholar
Evensen, G. 2009 Data Assimilation: The Ensemble Kalman Filter. Springer.10.1007/978-3-642-03711-5CrossRefGoogle Scholar
Fan, Y., Li, W., Atzori, M., Pozuelo, R., Schlatter, P. & Vinuesa, R. 2020 Decomposition of the mean friction drag in adverse-pressure-gradient turbulent boundary layers. Phys. Rev. Fluids 5 (11), 114608.10.1103/PhysRevFluids.5.114608CrossRefGoogle Scholar
Fang, Y., Zhao, Y., Waschkowski, F., Ooi, A.S.H. & Sandberg, R.D. 2023 Toward more general turbulence models via multicase computational-fluid-dynamics-driven training. AIAA J. 61 (5), 2100–2115.10.2514/1.J062572CrossRefGoogle Scholar
Gatski, T.B. & Jongen, T. 2000 Nonlinear eddy viscosity and algebraic stress models for solving complex turbulent flows. Prog. Aerosp. Sci. 36 (8), 655–682.10.1016/S0376-0421(00)00012-9CrossRefGoogle Scholar
Gritskevich, M.S., Garbaruk, A.V., Schütze, J. & Menter, F.R. 2012 Development of DDES and IDDES formulations for the k-
$\omega$
shear stress transport model. Flow Turbul. Combust. 88, 431–449.10.1007/s10494-011-9378-4CrossRefGoogle Scholar
$\omega$
shear stress transport model. Flow Turbul. Combust. 88, 431–449.10.1007/s10494-011-9378-4CrossRefGoogle ScholarHan, Y., He, Y. & Le, J. 2020 Modification to improved delayed detached-eddy simulation regarding the log-layer mismatch. AIAA J. 58 (2), 712–721.10.2514/1.J058535CrossRefGoogle Scholar
Harris, V.G., Graham, J.A.H. & Corrsin, S. 1977 Further experiments in nearly homogeneous turbulent shear flow. J. Fluid Mech. 81 (4), 657–687.10.1017/S0022112077002286CrossRefGoogle Scholar
Heinz, S. 2020 A review of hybrid RANS-LES methods for turbulent flows: concepts and applications. Prog. Aerosp. Sci. 114, 100597.10.1016/j.paerosci.2019.100597CrossRefGoogle Scholar
Jasak, H., et al. 2007 OpenFOAM: a C++ library for complex physics simulations. In International Workshop on Coupled Methods in Numerical Dynamics, pp. 1–20. IUC Dubrovnik Croatia.Google Scholar
Jee, S.K. & Shariff, K. 2014 Detached-eddy simulation based on the v2-f model. Intl J. Heat Fluid Flow 46, 84–101.10.1016/j.ijheatfluidflow.2013.12.006CrossRefGoogle Scholar
Jiang, C., Vinuesa, R., Chen, R., Mi, J., Laima, S. & Li, H. 2021 An interpretable framework of data-driven turbulence modeling using deep neural networks. Phys. Fluids 33 (5), 055133.10.1063/5.0048909CrossRefGoogle Scholar
Klein, M., Sadiki, A. & Janicka, J. 2003 A digital filter based generation of inflow data for spatially developing direct numerical or large eddy simulations. J. Comput. Phys. 186 (2), 652–665.10.1016/S0021-9991(03)00090-1CrossRefGoogle Scholar
Launder, B.E., Reece, G.J. & Rodi, W. 1975 Progress in the development of a Reynolds-stress turbulence closure. J. Fluid Mech. 68 (3), 537–566.10.1017/S0022112075001814CrossRefGoogle Scholar
Le Pape, A., Richez, F. & Deck, S. 2013 Zonal detached-eddy simulation of an airfoil in poststall condition. AIAA J. 51 (8), 1919–1931.10.2514/1.J052235CrossRefGoogle Scholar
Li, Q., Chen, X., Wang, G. & Liu, Y. 2022 A dynamic version of the improved delayed detached-eddy simulation based on the differential Reynolds-stress model. Phys. Fluids 34 (11), 115112.10.1063/5.0119552CrossRefGoogle Scholar
Lilly, D.K. 1991 A proposed modification of the Germano subgrid-scale closure method. Phys. Fluids A: Fluid Dyn. 4 (3), 633–635.10.1063/1.858280CrossRefGoogle Scholar
Ling, J., Kurzawski, A. & Templeton, J. 2016 Reynolds averaged turbulence modelling using deep neural networks with embedded invariance. J. Fluid Mech. 807, 155–166.10.1017/jfm.2016.615CrossRefGoogle Scholar
Ling, J. & Templeton, J. 2015 Evaluation of machine learning algorithms for prediction of regions of high Reynolds averaged Navier Stokes uncertainty. Phys. Fluids 27 (8), 085103.10.1063/1.4927765CrossRefGoogle Scholar
Liu, H., Yin, Z. & Liu, H. 2024 a Adaptive detached eddy simulation of turbulent combustion with the subgrid dissipation concept. Phys. Fluids 36 (5), 055154.10.1063/5.0208228CrossRefGoogle Scholar
Liu, H., Zhang, X., Yin, Z. & He, G. 2024 b An explicit algebraic stress-based delayed detached eddy simulation model for turbulent flows. AIAA J. 63 (6), 2249–2260.10.2514/1.J064894CrossRefGoogle Scholar
Liu, J., Chen, J. & Xiao, Z. 2024 c Improvements for the DDES model with consideration of grid aspect ratio and coherent vortex structures. Aerosp. Sci. Technol. 151, 109314.10.1016/j.ast.2024.109314CrossRefGoogle Scholar
Liu, Y., Zhou, Z.., Zhu, L. & Wang, S. 2021 Numerical investigation of flows around an axisymmetric body of revolution by using Reynolds-stress model based hybrid Reynolds-averaged Navier–Stokes/large eddy simulation. Phys. Fluids 33 (8), 085115.10.1063/5.0058016CrossRefGoogle Scholar
Lozano-Durán, A. & Bae, H.J. 2023 Machine learning building-block-flow wall model for large-eddy simulation. J. Fluid Mech. 963, A35.10.1017/jfm.2023.331CrossRefGoogle Scholar
Marstorp, L., Brethouwer, G., Grundestam, O. & Johansson, A.V. 2009 Explicit algebraic subgrid stress models with application to rotating channel flow. J. Fluid Mech. 639, 403–432.10.1017/S0022112009991054CrossRefGoogle Scholar
Matai, R. & Durbin, P.A. 2019 a Zonal eddy viscosity models based on machine learning. Flow Turbul. Combust. 103, 93–109.10.1007/s10494-019-00011-5CrossRefGoogle Scholar
Matai, R. & Durbin, P.A. 2019 b Large-eddy simulation of turbulent flow over a parametric set of bumps. J. Fluid Mech. 866, 503–525.10.1017/jfm.2019.80CrossRefGoogle Scholar
Michelén-Ströfer, C.A., Zhang, X.-L. & Xiao, H. 2021 DAFI: an open-source framework for ensemble-based data assimilation and field inversion. Commun. Comput. Phys. 29 (5), 1583–1622.10.4208/cicp.OA-2020-0178CrossRefGoogle Scholar
Park, J. & Choi, H. 2021 Toward neural-network-based large eddy simulation: application to turbulent channel flow. J. Fluid Mech. 914, A16.10.1017/jfm.2020.931CrossRefGoogle Scholar
Pettersson Reif, B.A. & Andersson, H.I. 2002 Prediction of turbulence-generated secondary mean flow in a square duct. Flow Turbul. Combust. 68, 41–61.10.1023/A:1015611721026CrossRefGoogle Scholar
Pirozzoli, S., Modesti, D., Orlandi, P. & Grasso, F. 2018 Turbulence and secondary motions in square duct flow. J. Fluid Mech. 840, 631–655.10.1017/jfm.2018.66CrossRefGoogle Scholar
Pont-Vílchez, A., Duben, A., Gorobets, A., Revell, A., Oliva, A. & Trias, F.X. 2021 New strategies for mitigating the gray area in delayed-detached eddy simulation models. AIAA J. 59 (9), 3331–3345.10.2514/1.J059666CrossRefGoogle Scholar
Pope, S.B. 1975 A more general effective-viscosity hypothesis. J. Fluid Mech. 72 (2), 331–340.10.1017/S0022112075003382CrossRefGoogle Scholar
Probst, A., Schwamborn, D., Garbaruk, A., Guseva, E., Shur, M., Strelets, M. & Travin, A. 2017 Evaluation of grey area mitigation tools within zonal and non-zonal RANS-LES approaches in flows with pressure induced separation. Intl J. Heat Fluid Flow 68, 237–247.10.1016/j.ijheatfluidflow.2017.08.008CrossRefGoogle Scholar
Reddy, K.R., Ryon, J.A. & Durbin, P.A. 2014 A DDES model with a Smagorinsky-type eddy viscosity formulation and log-layer mismatch correction. Intl J. Heat Fluid Flow 50, 103–113.10.1016/j.ijheatfluidflow.2014.06.002CrossRefGoogle Scholar
Renard, N. & Deck, S. 2016 A theoretical decomposition of mean skin friction generation into physical phenomena across the boundary layer. J. Fluid Mech. 790, 339–367.10.1017/jfm.2016.12CrossRefGoogle Scholar
Rodi, W. 1972 The prediction of free turbulent boundary layers by use of a two-equation model of turbulence. PhD Thesis, University of London, London.Google Scholar
Rotta, J.C. 1951 Statistische theorie nichthomogener turbulenz. Z. Phys. 129, 547–572.10.1007/BF01330059CrossRefGoogle Scholar
Sandberg, R.D. & Zhao, Y. 2022 Machine-learning for turbulence and heat-flux model development: a review of challenges associated with distinct physical phenomena and progress to date. Intl J. Heat Fluid Flow 95, 108983.10.1016/j.ijheatfluidflow.2022.108983CrossRefGoogle Scholar
Schmelzer, M., Dwight, R.P. & Cinnella, P. 2020 Discovery of algebraic Reynolds-stress models using sparse symbolic regression. Flow Turbul. Combust. 104 (2), 579–603.10.1007/s10494-019-00089-xCrossRefGoogle Scholar
Shur, M.L., Spalart, P.R., Strelets, M.K. & Travin, A.K. 2008 A hybrid RANS-LES approach with delayed-DES and wall-modelled LES capabilities. Intl J. Heat Fluid Flow 29 (6), 1638–1649.10.1016/j.ijheatfluidflow.2008.07.001CrossRefGoogle Scholar
Singh, A.P., Medida, S. & Duraisamy, K. 2017 Machine-learning-augmented predictive modeling of turbulent separated flows over airfoils. AIAA J. 55 (7), 2215–2227.10.2514/1.J055595CrossRefGoogle Scholar
Sirignano, J. & MacArt, J.F. 2023 Deep learning closure models for large-eddy simulation of flows around bluff bodies. J. Fluid Mech. 966, A26.10.1017/jfm.2023.446CrossRefGoogle Scholar
Sirignano, J., MacArt, J.F. & Freund, J.B. 2020 DPM: a deep learning PDE augmentation method with application to large-eddy simulation. J. Comput. Phys. 423, 109811.10.1016/j.jcp.2020.109811CrossRefGoogle Scholar
Spalart, P.R. 1997 Comments on the feasibility of LES for wings, and on a hybrid RANS/LES approach. In Proceedings of First AFOSR International Conference on DNS/LES. Greyden Press.Google Scholar
Spalart, P.R. 2000 Strategies for turbulence modelling and simulations. Intl J. Heat Fluid Flow 21 (3), 252–263.10.1016/S0142-727X(00)00007-2CrossRefGoogle Scholar
Spalart, P.R. 2009 Detached-eddy simulation. Annu. Rev. Fluid Mech. 41, 181–202.10.1146/annurev.fluid.010908.165130CrossRefGoogle Scholar
Spalart, P.R., Deck, S., Shur, M.L., Squires, K.D., Strelets, M.K. & Travin, A. 2006 A new version of detached-eddy simulation, resistant to ambiguous grid densities. Theor. Comput. Fluid Dyn. 20 (3), 181.10.1007/s00162-006-0015-0CrossRefGoogle Scholar
Taulbee, D.B. 1992 An improved algebraic Reynolds stress model and corresponding nonlinear stress model. Phys. Fluids A: Fluid Dyn. 4 (11), 2555–2561.10.1063/1.858442CrossRefGoogle Scholar
Volpiani, P.S. 2024 Are random forests better suited than neural networks to augment RANS turbulence models? Intl J. Heat Fluid Flow 107, 109348.10.1016/j.ijheatfluidflow.2024.109348CrossRefGoogle Scholar
Vreman, A.W. 2004 An eddy-viscosity subgrid-scale model for turbulent shear flow: algebraic theory and applications. Phys. Fluids 16 (10), 3670–3681.10.1063/1.1785131CrossRefGoogle Scholar
Wallin, S. & Johansson, A.V. 2000 An explicit algebraic Reynolds stress model for incompressible and compressible turbulent flows. J. Fluid Mech. 403, 89–132.10.1017/S0022112099007004CrossRefGoogle Scholar
Wang, G., Li, Q. & Liu, Y. 2021 IDDES method based on differential Reynolds-stress model and its application in bluff body turbulent flows. Aerosp. Sci. Technol. 119, 107207.10.1016/j.ast.2021.107207CrossRefGoogle Scholar
Wang, G., Tang, Y. & Liu, Y. 2025 Gray area mitigation in grid-adaptive simulation for wall-bounded turbulent flows. Intl J. Mech. Sci. 296, 110305.10.1016/j.ijmecsci.2025.110305CrossRefGoogle Scholar
Wang, J.-X., Wu, J.-L. & Xiao, H. 2017 Physics-informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data. Phys. Rev. Fluids 2 (3), 034603.10.1103/PhysRevFluids.2.034603CrossRefGoogle Scholar
Wu, C., Zhang, X.-L., Xu, D. & He, G. 2025 A framework for learning symbolic turbulence models from indirect observation data via neural networks and feature importance analysis. J. Comput. Phys. 537, 114068.10.1016/j.jcp.2025.114068CrossRefGoogle Scholar
Wu, J.-L., Xiao, H. & Paterson, E. 2018 Physics-informed machine learning approach for augmenting turbulence models: a comprehensive framework. Phys. Rev. Fluids 3 (7), 074602.10.1103/PhysRevFluids.3.074602CrossRefGoogle Scholar
Wu, X. 2017 Inflow turbulence generation methods. Annu. Rev. Fluid Mech. 49, 23–49.10.1146/annurev-fluid-010816-060322CrossRefGoogle Scholar
Xu, D., Wang, J., Yu, C. & Chen, S. 2023 Artificial-neural-network-based nonlinear algebraic models for large-eddy simulation of compressible wall-bounded turbulence. J. Fluid Mech. 960, A4.10.1017/jfm.2023.179CrossRefGoogle Scholar
Yin, Z. & Durbin, P.A. 2016 An adaptive DES model that allows wall-resolved eddy simulation. Intl J. Heat Fluid Flow 62, 499–509.10.1016/j.ijheatfluidflow.2016.08.009CrossRefGoogle Scholar
Yin, Z. & Durbin, P.A. 2022 Detached eddy simulation of transition in turbomachinery: linear compressor cascade. J. Turbomach. 144 (3), 031002.10.1115/1.4052309CrossRefGoogle Scholar
Yin, Z., Ge, X. & Durbin, P. 2021 Adaptive detached eddy simulation of transition under the influence of free-stream turbulence and pressure gradient. J. Fluid Mech. 915, A115.10.1017/jfm.2021.117CrossRefGoogle Scholar
Yin, Z., Reddy, K.R. & Durbin, P.A. 2015 On the dynamic computation of the model constant in delayed detached eddy simulation. Phys. Fluids 27 (2), 025105.10.1063/1.4907746CrossRefGoogle Scholar
Zhang, F., Zhou, Z., Yang, X. & He, G. 2025 Knowledge-integrated additive learning for consistent near-wall modelling of turbulent flows. J. Fluid Mech. 1011, R1.10.1017/jfm.2025.361CrossRefGoogle Scholar
Zhang, X.-L., Michelén-Ströfer, C. & Xiao, H. 2020 Regularized ensemble Kalman methods for inverse problems. J. Comput. Phys. 416, 109517.10.1016/j.jcp.2020.109517CrossRefGoogle Scholar
Zhang, X.-L., Xiao, H., He, G.-W. & Wang, S.Z. 2021 Assimilation of disparate data for enhanced reconstruction of turbulent mean flows. Comput. Fluids 224, 104962.10.1016/j.compfluid.2021.104962CrossRefGoogle Scholar
Zhang, X.-L., Xiao, H., Jee, S. & He, G. 2023 a Physical interpretation of neural network-based nonlinear eddy viscosity models. Aerosp. Sci. Technol. 142, 108632.10.1016/j.ast.2023.108632CrossRefGoogle Scholar
Zhang, X.-L., Xiao, H., Luo, X. & He, G. 2022 Ensemble Kalman method for learning turbulence models from indirect observation data. J. Fluid Mech. 949, A26.10.1017/jfm.2022.744CrossRefGoogle Scholar
Zhang, X.-L., Xiao, H., Luo, X. & He, G. 2023 b Combining direct and indirect sparse data for learning generalizable turbulence models. J. Comput. Phys. 489, 112272.10.1016/j.jcp.2023.112272CrossRefGoogle Scholar
Zhou, Z., He, G. & Yang, X. 2021 Wall model based on neural networks for LES of turbulent flows over periodic hills. Phys. Rev. Fluids 6 (5), 054610.10.1103/PhysRevFluids.6.054610CrossRefGoogle Scholar
Zhou, Z., Zhang, X.-L., He, G.-W. & Yang, X. 2025 A wall model for separated flows: embedded learning to improve a posteriori performance. J. Fluid Mech. 1002, A3.10.1017/jfm.2024.1127CrossRefGoogle Scholar
Zhuchkov, R.N. & Utkina, A.A. 2016 Combining the SSG/LRR-
$\omega$
differential Reynolds stress model with the detached eddy and laminar-turbulent transition models. Fluid Dyn. 51 (6), 733–744.10.1134/S001546281606003XCrossRefGoogle Scholar
$\omega$
differential Reynolds stress model with the detached eddy and laminar-turbulent transition models. Fluid Dyn. 51 (6), 733–744.10.1134/S001546281606003XCrossRefGoogle Scholar