Abstract
A data-driven wall function estimation approach is proposed, aimed at accounting for non-equilibrium effects in turbulent boundary layers in RANS simulations of wall bounded flows. While keeping key simplifying hypothesis of standard wall functions and their general structure, the law-of-the-wall is replaced by a fully connected feed-forward neural network. The latter is trained to infer wall friction from the local flow state at the first of-wall nodes, described by an extended set of flow variables and gradients. For this purpose, the neural network is trained on high-fidelity wall resolved simulation data. It is then applied to formulate two different wall functions trained on high-fidelity data: a backward-facing step and a round jet impacting a flat wall. After integration into an industrial CFD code, they are applied to perform RANS simulations of the flow configurations they were trained for, and are shown to yield a largely improved prediction of wall friction as compared to standard wall functions. Finally, key issues related to the practical usability in RANS applications of the proposed data-driven approach are critically discussed.
Similar content being viewed by others
Data Availability
Data available on request from the authors.
References
Abadi, M., et al.: TensorFlow: large-scale machine learning on heterogeneous systems (2015). https://www.tensorflow.org/, software available from tensorflow.org
Amarloo, A., Cinnella, P., Iosifidis, A., et al.: Data-driven Reynolds stress models based on the frozen treatment of Reynolds stress tensor and Reynolds force vector. Phys. Fluids 35(7), 075154 (2023). https://doi.org/10.1063/5.0160977
Aubagnac-Karkar, D., Mehl, C.: NNICE: neural network inference in C made easy (2023). https://doi.org/10.5281/zenodo.7645515, https://github.com/aubagnacd/NNICE
Beck, A., Flad, D., Munz, C.D.: Deep neural networks for data-driven les closure models. J. Comput. Phys. 398, 108910 (2019)
Billard, F., Laurence, D., Osman, K.: Adaptive wall functions for an elliptic blending eddy viscosity model applicable to any mesh topology. Flow Turbul. Combust. 94, 817–842 (2015). https://doi.org/10.1007/s10494-015-9600-x
Bin, Y., Park, G.I., Lv, Y., Yang, X.I.A: Large-Eddy Simulation of Separated Flows on Unconventionally Coarse Grids. Proceedings of the ASME 2023 International Mechanical Engineering Congress and Exposition. Volume 9: Fluids Engineering. New Orleans, Louisiana, USA. October 29–November 2, 2023. V009T10A018. ASME. https://doi.org/10.1115/IMECE2023-116879
Brenner, M.P., Eldredge, J.D., Freund, J.B.: Perspective on machine learning for advancing fluid mechanics. Phys. Rev. Fluids 4(10), 1 (2019)
Brunton, S.L., Noack, B.R., Koumoutsakos, P.: Machine learning for fluid mechanics. Annu. Rev. Fluid Mech. 52(1), 477–508 (2020)
Calzolari, G., Liu, W.: Deep learning to replace, improve, or aid CFD analysis in built environment applications: a review. Build. Environ. 206, 108315 (2021)
Cheng, C., Zhang, G.T.: Deep learning method based on physics informed neural network with resnet block for solving fluid flow problems. Water 13, 25 (2021)
Choi, H., Moin, P.: Grid-point requirements for large eddy simulation: Chapman’s estimates revisited. Phys. Fluids 24(1), 011702 (2012)
Craft, T.J., Gant, S.E., Gerasimov, A.V., et al.: Development and application of wall-function treatments for turbulent forced and mixed convection flows. Fluid Dyn. Res. 38(2–3), 127 (2006). https://doi.org/10.1016/j.fluiddyn.2004.11.002
Duraisamy, K.: Perspectives on machine learning-augmented Reynolds-averaged and large eddy simulation models of turbulence. Phys. Rev. Fluids 6(5), 050504 (2021). https://doi.org/10.1103/PhysRevFluids.6.050504
Duraisamy, K., Iaccarino, G., Xiao, H.: Turbulence modeling in the age of data. Annu. Rev. Fluid Mech. 51(1), 357–377 (2019)
Ferdian, E., Suinesiaputra, A., Dubowitz, D.J., et al.: 4DFlowNet: super-resolution 4D flow MRI using deep learning and computational fluid dynamics. Front. Phys. 8, 486 (2020)
Grenouilloux, A., Balarac, G., Leparoux, J., et al.: On the use of kinetic-energy balance for the feature-based mesh adaptation applied to large eddy simulation in complex geometries. In: Proceedings ASME Turbo Expo (2022)
Grenouilloux, A., Moureau, V., Lartigue, G., et al.: Feature-based mesh adaptation applied to the large eddy simulation of multiple jets impinging on a surface. In: Proceedings UK Heat Transfer Conference (2021)
Güemes, A., Discetti, S., Ianiro, A., et al.: From coarse wall measurements to turbulent velocity fields through deep learning. Phys. Fluids 33(7), 075121 (2021)
Guennebaud, G., Jacob, B., et al.: Eigen v3 (2010). http://eigen.tuxfamily.org
Guo, M., Hesthaven, J.S.: Reduced order modeling for nonlinear structural analysis using gaussian process regression. Comput. Methods Appl. Mech. Eng. 341, 807–826 (2018)
Guo, M., Hesthaven, J.S.: Data-driven reduced order modeling for time-dependent problems. Comput. Methods Appl. Mech. Eng. 345, 75–99 (2019)
Hasegawa, K., Fukami, K., Murata, T., et al.: Machine-learning-based reduced-order modeling for unsteady flows around bluff bodies of various shapes. Theoret. Comput. Fluid Dyn. 55(12), 4013 (2020)
Huang, X.L.D., Yang, X.I.A., Kunz, R.F.: Wall-modeled large eddy simulation of spanwise rotating turbulent channels-comparing a physics-based approach and a data-based approach. Phys. Fluids 31(12), 125105 (2019)
Jiang, C., Mi, J., Laima, S., et al.: A novel algebraic stress model with machine-learning-assisted parameterization. Energies 13(1), 258 (2020)
Joshi, A., Assam, A., Nived, M.R., et al.: A generalised wall function including compressibility and pressure-gradient terms for the Spalart–Allmaras turbulence model. J. Turbul. 20(10), 626–660 (2019)
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. In: 3rd International Conference for Learning Representations, San Diego, p. 15 (2015)
Launder, B.E., Rodi, W.: The turbulent wall jet measurements and modeling. Annu. Rev. Fluid Mech. 15(1), 429–459 (1983). https://doi.org/10.1146/annurev.fl.15.010183.002241
Launder, B., Spalding, D.: The numerical computation of turbulent flows. Comput. Methods Appl. Mech. Eng. 3(2), 269–289 (1974). https://doi.org/10.1016/0045-7825(74)90029-2
Ling, J., Jones, R., Templeton, J.: Machine learning strategies for systems with invariance properties. J. Comput. Phys. 318, 22–35 (2016)
Ling, J., Kurzawski, A., Templeton, J.: Reynolds averaged turbulence modelling using deep neural networks with embedded invariance. J. Fluid Mech. 807, 155–166 (2016)
Lozano-Durán, A., Giometto, M.G., Park, G.I., et al.: Non-equilibrium three-dimensional boundary layers at moderate Reynolds numbers. J. Fluid Mech. 883, A20 (2020). https://doi.org/10.1017/jfm.2019.869
Matai, R., Durbin, P.A.: Zonal eddy viscosity models based on machine learning. Flow Turbul. Combust. 103(1), 93–109 (2019)
Mehl, C., Aubagnac-Karkar, D.: On-the-fly accuracy evaluation of artificial neural networks and hybrid method to improve the robustness of neural network accelerated chemistry solving. Phys. Fluids. 10(1063/5), 0151026 (2023)
Menter, F., Kuntz, M., Langtry, R.: Ten years of industrial experience with the SST turbulence model. Turbulence Heat Mass Transf. 4, 625–632 (2003)
Milano, M., Koumoutsakos, P.: Neural network modeling for near wall turbulent flow. J. Comput. Phys. 182(1), 1–26 (2002)
Ong, K.C., Chan, A.: A unified wall function for compressible turbulence modelling. J. Turbul. 19(5), 414–430 (2018)
Parish, E., Duraisamy, K.: A paradigm for data-driven predictive modeling using field inversion and machine learning. J. Comput. Phys. 305, 758–774 (2015)
Pedregosa, F., Varoquaux, G., Gramfort, A., et al.: Scikit-learn: machine learning in python. J. Mach. Learn. Res. 12, 2825–2830 (2011)
Pope, SB.: Turbulent flows (2000)
Popovac, M., Hanjalic, K.: Compound wall treatment for RANS computation of complex turbulent flows and heat transfer. Flow Turbul. Combust. 78(2), 177–202 (2007)
Richards, K., Senecal, P., Pomraning, E.: CONVERGE 3.0.19 (2022)
Rudy, S.H., Brunton, S.L., Proctor, J.L., et al.: Data-driven discovery of partial differential equations. Sci. Adv. 3(4), e1602614 (2017)
Saïdi, I.B.H., Schmelzer, M., Cinnella, P., et al.: CFD-driven symbolic identification of algebraic Reynolds-stress models. J. Comput. Phys. 457, 111037 (2022). https://doi.org/10.1016/j.jcp.2022.111037
Schmelzer, M., Dwight, R.P., Cinnella, P.: Discovery of algebraic Reynolds-stress models using sparse symbolic regression. Flow Turbul. Combust. 104, 579–603 (2020)
Shih, T.H., Povinelli, L.A., Liu, N.S.: Application of generalized wall function for complex turbulent flows. In: Rodi, W., Fueyo, N. (eds.) Engineering Turbulence Modelling and Experiments, vol. 5, pp. 177–186. Elsevier, Amsterdam (2002)
Slotnick, J., Khodadoust, A., Alonso, J., et al.: CFD vision 2030 study: a path to revolutionary computational aerosciences (2014). https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20140003093.pdf
Tieghi, L., Corsini, A., Delibra, G., et al.: Assessment of a machine-learnt adaptive wall-function in a compressor cascade with sinusoidal leading edge. J. Eng. Gas Turbines Power 8, 142 (2020)
Tieghi, L., Corsini, A., Delibra, G., et al.: A machine-learnt wall function for rotating diffusers. J. Turbomach. 143(8), 081012 (2021)
Vaddireddy, H., Rasheed, A., Staples, A.E., et al.: Feature engineering and symbolic regression methods for detecting hidden physics from sparse sensor observation data. Phys. Fluids 32, 1 (2020)
Wang, J.X., Wu, J.L., Xiao, H.: Physics-informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data. Phys. Rev. Fluids 2(3), 034603 (2017)
Wang, H., Yang, Z., Li, B., et al.: Predicting the near-wall velocity of wall turbulence using a neural network for particle image velocimetry. Phys. Fluids 32(11), 115105 (2020)
Weatheritt, J., Sandberg, R.: A novel evolutionary algorithm applied to algebraic modifications of the RANS stress–strain relationship. J. Comput. Phys. 325, 22–37 (2016)
Wilcox, D.: Turbulence Modeling for CFD, 3rd edn (2006)
Wu, J.L., Wang, J.X., Xiao, H.: A bayesian calibration-prediction method for reducing model-form uncertainties with application in RANS simulations. Flow Turbul. Combust. 97(3), 761–786 (2016)
Wu, J.L., Xiao, H., Paterson, E.: Physics-informed machine learning approach for augmenting turbulence models: a comprehensive framework. Phys. Rev. Fluids 3(7), 18 (2018)
Xie, C., Wang, J., Li, H., et al.: Spatially multi-scale artificial neural network model for large eddy simulation of compressible isotropic turbulence. AIP Adv. 10(1), 015044 (2020)
Yang, X.I.A., Zafar, S., Wang, J.X., et al.: Predictive large eddy simulation wall modeling via physics-informed neural networks. Phys. Rev. Fluids 4, 034602 (2019)
Zhou, Z., He, G., Wang, S., et al.: Subgrid-scale model for large eddy simulation of isotropic turbulent flows using an artificial neural network. Comput. Fluids 195, 104319 (2019)
Zhou, Z., He, G., Yang, X.: Wall model based on neural networks for les of turbulent flows over periodic hills. Phys. Rev. Fluids 6, 054610 (2021)
Funding
N/A.
Author information
Authors and Affiliations
Contributions
E.R. developed the methodology, performed the computations and wrote the original draft of the manuscript text. A.P., C.A., M.M.Z and R.P. contributed to the development of the methodology, the interpretation of the results and the drafting of the manuscript. D.A.K. contributed to the technical implementation of the methodology. All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors declared that they have no conflict of interest.
Ethical Approval
N/A.
Informed Consent
N/A.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix: Formulation of the LWWF
The LWWF is based on the Launder and Spalding wall model (Launder and Spalding 1974). The balance between production and dissipation rates of turbulent kinetic energy in the wall bounbary layer leads to the following estimation of the friction velocity:
with \(C_{\mu }=0.09\) and \(k_1\) the value of the turbulent kinetic energy at the centroid of the first off-wall cell p. This friction velocity \(u_{k}\) is used to compute the dimensionless wall distance \(h^{+}\):
with h the wall normal distance to the centroid of the first off-wall cell. Combining the logarithmic law-of-the-wall with Eq. A2 yields a second estimation of the friction velocity \(u_{\tau }\):
with \(E=e^{\kappa B}\), \(\kappa = 0.4187\) and \(B=5.5\). The wall shear stress is then computed as follows:
Regarding the wall boundary conditions applied to the turbulence quantities, a zero-gradient Neumann condition is used for k and a wall law for the specific dissipation rate \(\omega\):
Appendix B: Filtering High-Fidelity Data to Coarse RANS Meshes
A two-dimensional filtering along the local normal distance to the wall is applied to HiFi data in order to map the mean flow variables from the highly resolved mesh having served to generate it, to the coarser mesh resolution typical of practical RANS simulations.
As illustrated in Fig. 14, the unfiltered value \(I^{hf}(y_i)\) of any mean flow variable I in the initial HiFi data is replaced by an area-averaged value \(I^{filt}(y_i)\) over all cells of the HiFi mesh intersecting with the coarse RANS cell of area \(S_{filt} = y_i ^2\), as:
where \(N^{hf}\) is the number of cells in the HiFi mesh.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Rondeaux, E., Poubeau, A., Angelberger, C. et al. Exploring the Potential and the Practical Usability of a Machine Learning Approach for Improving Wall Friction Predictions of RANS Wall Functions in Non-equilibrium Turbulent Flows. Flow Turbulence Combust 112, 975–1000 (2024). https://doi.org/10.1007/s10494-024-00539-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10494-024-00539-1