Evaluation of machine learning algorithms for prediction of regions of high Reynolds averaged Navier Stokes uncertainty

Reynolds Averaged Navier Stokes (RANS) models are widely used in industry to predict fluid flows, despite their acknowledged deficiencies. Not only do RANS models often produce inaccurate flow predictions, but there are very limited diagnostics available to assess RANS accuracy for a given flow configuration. If experimental or higher fidelity simulation results are not available for RANS validation, there is no reliable method to evaluate RANS accuracy. This paper explores the potential of utilizing machine learning algorithms to identify regions of high RANS uncertainty. Three different machine learning algorithms were evaluated: support vector machines, Adaboost decision trees, and random forests. The algorithms were trained on a database of canonical flow configurations for which validated direct numerical simulation or large eddy simulation results were available, and were used to classify RANS results on a point-by-point basis as having either high or low uncertainty, based on the breakdown of specific RANS modeling assumptions. Classifiers were developed for three different basic RANS eddy viscosity model assumptions: the isotropy of the eddy viscosity, the linearity of the Boussinesq hypothesis, and the non-negativity of the eddy viscosity. It is shown that these classifiers are able to generalize to flows substantially different from those on which they were trained. Feature selection techniques, model evaluation, and extrapolation detection are discussed in the context of turbulence modeling applications.

[1]  P. Mahalanobis On the generalized distance in statistics , 1936 .

[2]  B. Launder,et al.  Progress in the development of a Reynolds-stress turbulence closure , 1975, Journal of Fluid Mechanics.

[3]  S. Pope A more general effective-viscosity hypothesis , 1975, Journal of Fluid Mechanics.

[4]  W. Rodi A new algebraic relation for calculating the Reynolds stresses , 1976 .

[5]  Akira Yoshizawa,et al.  Turbulent channel and Couette flows using an anisotropic k-epsilon model , 1987 .

[6]  T. Gatski,et al.  On explicit algebraic stress models for complex turbulent flows , 1992, Journal of Fluid Mechanics.

[7]  Kevin W. Bowyer,et al.  Combination of multiple classifiers using local accuracy estimates , 1996, Proceedings CVPR IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[8]  Michel Deville,et al.  On predicting the turbulence‐induced secondary flows using nonlinear k‐ε models , 1996 .

[9]  B. Launder,et al.  Development and application of a cubic eddy-viscosity model of turbulence , 1996 .

[10]  Christopher M. Bishop,et al.  Classification and regression , 1997 .

[11]  Thorsten Joachims,et al.  Making large scale SVM learning practical , 1998 .

[12]  John Kim,et al.  DIRECT NUMERICAL SIMULATION OF TURBULENT CHANNEL FLOWS UP TO RE=590 , 1999 .

[13]  Sayan Mukherjee,et al.  Feature Selection for SVMs , 2000, NIPS.

[14]  S. Pope Turbulent Flows: FUNDAMENTALS , 2000 .

[15]  Arne V. Johansson,et al.  An explicit algebraic Reynolds stress model for incompressible and compressible turbulent flows , 2000, Journal of Fluid Mechanics.

[16]  Nitesh V. Chawla,et al.  Creating ensembles of classifiers , 2001, Proceedings 2001 IEEE International Conference on Data Mining.

[17]  Nitesh V. Chawla,et al.  Distributed Pasting of Small Votes , 2002, Multiple Classifier Systems.

[18]  Stefan P. Domino,et al.  SIERRA/Fuego: A Multi-Mechanics Fire Environment Simulation Tool , 2003 .

[19]  Robert P. Sheridan,et al.  Random Forest: A Classification and Regression Tool for Compound Classification and QSAR Modeling , 2003, J. Chem. Inf. Comput. Sci..

[20]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[21]  Lawrence O. Hall,et al.  A Comparison of Ensemble Creation Techniques , 2004, Multiple Classifier Systems.

[22]  Giles Hooker Diagnosing extrapolation: tree-based density estimation , 2004, KDD '04.

[23]  Ivan Bermejo-Moreno,et al.  On the non-local geometry of turbulence , 2008, Journal of Fluid Mechanics.

[24]  Gianluca Iaccarino,et al.  A numerical study of scalar dispersion downstream of a wall-mounted cube using direct simulations and algebraic flux models , 2010 .

[25]  Alfredo Pinelli,et al.  Reynolds number dependence of mean flow structure in square duct turbulence , 2010, Journal of Fluid Mechanics.

[26]  Riccardo Rossi,et al.  A numerical study of algebraic flux models for heat and mass transport simulation in complex flows , 2010 .

[27]  Qiqi Wang,et al.  Uncertainty Quantification of Structural Uncertainties in RANS Simulations of Complex Flows , 2011 .

[28]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[29]  Todd A. Oliver,et al.  Bayesian uncertainty quantification applied to RANS turbulence models , 2011 .

[30]  Sai Hung Cheung,et al.  Bayesian uncertainty analysis with applications to turbulence modeling , 2011, Reliab. Eng. Syst. Saf..

[31]  Qiqi Wang,et al.  Quantification of Structural Uncertainties in the k -w Turbulence Model , 2011 .

[32]  Gianluca Iaccarino,et al.  RANS modeling of turbulent mixing for a jet in supersonic cross flow: model evaluation and uncertainty quantification , 2012 .

[33]  Yoav Freund,et al.  Boosting: Foundations and Algorithms , 2012 .

[34]  Brendan D. Tracey,et al.  Application of supervised learning to quantify uncertainties in turbulence and combustion modeling , 2013 .

[35]  Srinivasan Arunajatesan,et al.  Validation of an FSI Modeling Framework for Internal Captive Carriage Applications. , 2013 .

[36]  Gianluca Iaccarino,et al.  Numerical analysis and modeling of plume meandering in passive scalar dispersion downstream of a wall-mounted cube , 2013 .

[37]  Christopher J. Elkins,et al.  Turbulent transport in an inclined jet in crossflow , 2013 .

[38]  S. Arunajatesan,et al.  Tuning a RANS k-e model for jet-in-crossflow simulations. , 2013 .

[39]  Hester Bijl,et al.  Bayesian estimates of parameter variability in the k-ε turbulence model , 2014, J. Comput. Phys..

[40]  R. D. Radenkovic,et al.  Anisotropy analysis of turbulent swirl flow , 2014 .

[41]  Srinivasan Arunajatesan,et al.  Estimation of k-ε parameters using surrogate models and jet-in-crossflow data , 2014 .

[42]  Srinivasan Arunajatesan,et al.  Bayesian calibration of a k-e turbulence model for predictive jet-in-crossflow simulations. , 2014 .

[43]  Julia Ling,et al.  Near Wall Modeling for Trailing Edge Slot Film Cooling , 2015 .

[44]  Brendan D. Tracey,et al.  A Machine Learning Strategy to Assist Turbulence Model Development , 2015 .

[45]  Guilhem Lacaze,et al.  Flow topologies and turbulence scales in a jet-in-cross-flow , 2015 .

[46]  Anand Pratap Singh,et al.  New Approaches in Turbulence and Transition Modeling Using Data-driven Techniques , 2015 .

[47]  Julia Ling,et al.  Analysis of Turbulent Scalar Flux Models for a Discrete Hole Film Cooling Flow , 2015 .