Prediction of Reynolds stresses in high-Mach-number turbulent boundary layers using physics-informed machine learning
Modeled Reynolds stress is a major source of model-form uncertainties in Reynolds-averaged Navier–Stokes (RANS) simulations. Recently, a physics-informed machine learning (PIML) approach has been proposed for reconstructing the discrepancies in RANS-modeled Reynolds stresses. The merits of the PIML framework have been demonstrated in several canonical incompressible flows. However, its performance on high-Mach-number flows is still not clear. In this work, we use the PIML approach to predict the discrepancies in RANS-modeled Reynolds stresses in high-Mach-number flat-plate turbulent boundary layers by using an existing DNS database. Specifically, the discrepancy function is first constructed using a DNS training flow and then used to correct RANS-predicted Reynolds stresses under flow conditions different from the DNS. The machine learning technique is shown to significantly improve RANS-modeled turbulent normal stresses, the turbulent kinetic energy, and the Reynolds stress anisotropy. Improvements are consistently observed when different training datasets are used. Moreover, a high-dimensional visualization technique and a distance metrics are used to provide a priori assessment of prediction confidence based only on RANS simulations. This study demonstrates that the PIML approach is a computationally affordable technique for improving the accuracy of RANS-modeled Reynolds stresses for high-Mach-number turbulent flows when there is a lack of experiments and high-fidelity simulations.
KeywordsData-driven Reynolds-averaged Navier–Stokes High-speed flow Direct numerical simulation
Unable to display preview. Download preview PDF.
The DNS database was produced based upon the work supported by AFOSR under Grant FA9550-14-1-0170 (Program Manager I. Leyva) and NASA Langley Research Center under Grant NNL09AA00A (through the National Institute of Aerospace). Computational resources for the DNS were provided by the NASA Advanced Supercomputing Division, the DoD High-Performance Computing Modernization Program, and the NSF’s Petascale Computing Resource Allocations Program (NSF ACI-1640865). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the United States Air Force. We also thank the anonymous reviewers for their comments, which helped improving the quality and clarity of the manuscript.
- 1.Ansys fluent user guide, release 15.0. ANSYS Inc (2013)Google Scholar
- 5.Dow, E., Wang, Q.: Quantification of structural uncertainties in the \(k\)–\(\omega \) turbulence model. In: 52nd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics and Materials Conference, AIAA, Denver, Colorado, pp. 2011–1762 (2011)Google Scholar
- 6.Duan, L., Choudhari, M.M.: Analysis of numerical simulation database for pressure fluctuations induced by high-speed turbulent boundary layers. AIAA Paper 2014-2912 (2014)Google Scholar
- 10.Duraisamy, K., Iaccarino, G., Xiao, H.: Turbulence modeling in the age of data. (2018). arXiv preprint arXiV:1804.00183
- 15.Friedman, J., Hastie, T., Tibshirani, R.: The elements of statistical learning, vol 1. Springer series in statistics New York, NY, USA (2001)Google Scholar
- 23.Keyes, F.G.: A summary of viscosity and heat-conduction data for He, A, \(H_2\), \(O_2\), \(CO\), \(CO_2\), \(H_2 O\), and air. Trans. Am. Soc. Mech. Eng. 73, 589–596 (1951)Google Scholar
- 31.Oliver, T.A., Moser, R.D.: Bayesian uncertainty quantification applied to RANS turbulence models. In: Journal of Physics: Conference Series, IOP Publishing vol 318, p. 042032 (2011)Google Scholar
- 33.Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M., Duchesnay, E.: Scikit-learn: machine learning in Python. J. Mach. Learn. Res. 12, 2825–2830 (2011)MathSciNetzbMATHGoogle Scholar
- 37.Singh, A.P., Duraisamy, K.: Using field inversion to quantify functional errors in turbulence closures. Phys. Fluids 28(045), 110 (2016)Google Scholar
- 38.Smits, A.J., Dussauge, J.P.: Turbulent Shear Layers in Supersonic Flow, 2nd edn. American Institute of Physics, College Park (2006)Google Scholar
- 42.Touber, E., Sandham, N.D.: Oblique shock impinging on a turbulent boundary layer: low-frequency mechanisms. AIAA Paper 2008-4170 (2008)Google Scholar
- 44.Wang, J., Wu, J., Ling, J., Iaccarino, G., Xiao, H.: Physics-informed machine learning for predictive turbulence modeling: toward a complete framework. In: 2016 Proceedings of the CTR Summer Program, Stanford University, CA, p. 1 (2016a)Google Scholar
- 47.Wang, J.X., Wu, J., Ling, J., Iaccarino, G., Xiao, H.: A comprehensive physics-informed machine learning framework for predictive turbulence modeling, submitted, (2017a). arXiv:1701.07102
- 48.Wang, J.X., Wu, J., Xiao, H.: Physics informed machine learning approach for reconstructing Reynolds stress modeling discrepancies based on DNS data. Phys. Rev. Fluids 2(3), 1–22 (2017b)Google Scholar
- 50.Wu, J., Wang, J., Xiao, H., Ling, J.: Visualization of high dimensional turbulence simulation data using t-SNE. In: 19th AIAA Non-Deterministic Approaches Conference, p. 1770 (2017a)Google Scholar
- 51.Wu, J.L., Wang, J.X., Xiao, H.: A Bayesian calibration–prediction method for reducing model-form uncertainties with application in RANS simulations. Flow, Turbulence and Combustion, pp. 1–26 (2015)Google Scholar
- 52.Wu, J.L., Wang, J.X., Xiao, H., Ling, J.: A priori assessment of prediction confidence for data-driven turbulence modeling. Flow, Turbulence and Combustion, pp. 1–22 (2017b)Google Scholar