Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier-Stokes simulations: a data-driven, physics-informed Bayesian approach
From MaRDI portal
Publication:525932
DOI10.1016/J.JCP.2016.07.038zbMATH Open1371.76082arXiv1508.06315OpenAlexW2490045648MaRDI QIDQ525932FDOQ525932
Authors: Heng Xiao, Jin-Long Wu, Jian-Xun Wang, Rui Sun, Christopher J. Roy 
Publication date: 8 May 2017
Published in: Journal of Computational Physics (Search for Journal in Brave)
Abstract: Despite their well-known limitations, Reynolds-Averaged Navier-Stokes (RANS) models are still the workhorse tools for turbulent flow simulations in today's engineering application. For many practical flows, the turbulence models are by far the largest source of uncertainty. In this work we develop an open-box, physics-informed Bayesian framework for quantifying model-form uncertainties in RANS simulations. Uncertainties are introduced directly to the Reynolds stresses and are represented with compact parameterization accounting for empirical prior knowledge and physical constraints (e.g., realizability, smoothness, and symmetry). An iterative ensemble Kalman method is used to assimilate the prior knowledge and observation data in a Bayesian framework, and to propagate them to posterior distributions of velocities and other Quantities of Interest (QoIs). We use two representative cases, the flow over periodic hills and the flow in a square duct, to evaluate the performance of the proposed framework. Simulation results suggest that, even with very sparse observations, the posterior mean velocities and other QoIs have significantly better agreement with the benchmark data compared to the baseline results. At most locations the posterior distribution adequately captures the true model error within the developed model form uncertainty bounds. The framework is a major improvement over existing black-box, physics-neutral methods for model-form uncertainty quantification, where prior knowledge and details of the models are not exploited. This approach has potential implications in many fields in which the governing equations are well understood but the model uncertainty comes from unresolved physical processes.
Full work available at URL: https://arxiv.org/abs/1508.06315
Recommendations
- Quantifying model form uncertainty in Reynolds-averaged turbulence models with Bayesian deep neural networks
- Bayesian uncertainty quantification of turbulence models based on high-order adjoint
- Evaluation of physics constrained data-driven methods for turbulence model uncertainty quantification
- Physics-informed covariance kernel for model-form uncertainty quantification with application to turbulent flows
- A sparse grid method for Bayesian uncertainty quantification with application to large eddy simulation turbulence models
- Bayesian assessment of uncertainty in viscosity closure models for turbidity currents computations
- Efficient prediction of turbulent flow quantities using a Bayesian hierarchical multifidelity model
- Uncertainty quantification for data assimilation in a steady incompressible Navier-Stokes problem
- A Bayesian approach to physics-based reconstruction of incompressible flows
uncertainty quantificationturbulence modelingReynolds-averaged Navier-Stokes equationsensemble Kalman filtering
Cites Work
- Bayesian calibration of computer models. (With discussion)
- Spatiotemporal Hierarchical Bayesian Modeling Tropical Ocean Surface Winds
- Orthonormal bases of compactly supported wavelets
- Combining Field Data and Computer Simulations for Calibration and Prediction
- A calculation procedure for heat, mass and momentum transfer in three- dimensional parabolic flows
- Numerical study of the turbulent flow past an airfoil with trailing edge separation
- Direct numerical simulation of turbulent flow in a square duct
- Radial Basis Functions
- Data Assimilation
- Spectral Methods for Uncertainty Quantification
- Title not available (Why is that?)
- Learning about physical parameters: the importance of model discrepancy
- Flow over periodic hills -- numerical and experimental study in a wide range of Reynolds numbers
- Title not available (Why is that?)
- Ensemble Kalman methods for inverse problems
- Calculation of turbulence-driven secondary motion in non-circular ducts
- A comprehensive framework for verification, validation, and uncertainty quantification in scientific computing
- Progress in the development of a Reynolds-stress turbulence closure
- A non-stationary covariance-based Kriging method for metamodelling in engineering design
- Maximum likelihood Bayesian averaging of uncertain model predictions
- Active subspace methods in theory and practice: applications to kriging surfaces
- Likelihood-informed dimension reduction for nonlinear inverse problems
- A paradigm for data-driven predictive modeling using field inversion and machine learning
- An inverse transformation for quadrilateral isoparameteric elements: Analysis and application
- Presentation of anisotropy properties of turbulence, invariants versus eigenvalue approaches
- Gaussian process emulation of dynamic computer codes
- The formation of streamwise vorticity in turbulent flow
Cited In (60)
- A probabilistic, data-driven closure model for RANS simulations with aleatoric, model uncertainty
- Data-driven approach for modeling Reynolds stress tensor with invariance preservation
- Bayesian interface technique-based inverse estimation of closure coefficients of standard \(k-\epsilon\) turbulence model by limiting the number of DNS data points for flow over a periodic hill
- Learning about structural errors in models of complex dynamical systems
- Optimal sensor placement for ensemble-based data assimilation using gradient-weighted class activation mapping
- A hybrid approach combining DNS and RANS simulations to quantify uncertainties in turbulence modelling
- Multi-fidelity modeling framework for nonlinear unsteady aerodynamics of airfoils
- Enforcing boundary conditions on physical fields in Bayesian inversion
- Performance assessment of the maximum likelihood ensemble filter and the ensemble Kalman filters for nonlinear problems
- Evaluation of physics constrained data-driven methods for turbulence model uncertainty quantification
- An efficient approach for quantifying parameter uncertainty in the SST turbulence model
- Real-time thermoacoustic data assimilation
- An efficient Bayesian uncertainty quantification approach with application to \(k\)-\(\omega\)-\(\gamma\) transition modeling
- Combined state and parameter estimation in level-set methods
- Turbulence modeling in the age of data
- A data-driven shock capturing approach for discontinuous Galerkin methods
- Bayesian uncertainty quantification of turbulence models based on high-order adjoint
- An adaptive minimum spanning tree multielement method for uncertainty quantification of smooth and discontinuous responses
- Model reduction with memory and the machine learning of dynamical systems
- Sharp interface approaches and deep learning techniques for multiphase flows
- A multigrid/ensemble Kalman filter strategy for assimilation of unsteady flows
- Bayesian assessment of uncertainty in viscosity closure models for turbidity currents computations
- Stochastic turbulence modeling in RANS simulations via multilevel Monte Carlo
- \(S\)-frame discrepancy correction models for data-informed Reynolds stress closure
- Quantifying turbulence model uncertainty in Reynolds-averaged Navier-Stokes simulations of a pin-fin array. I: Flow field
- Quantifying turbulence model uncertainty in Reynolds-averaged Navier-Stokes simulations of a pin-fin array. II: Scalar transport
- Bayesian calibration of computer models based on Takagi-Sugeno fuzzy models
- A random matrix approach for quantifying model-form uncertainties in turbulence modeling
- Quantifying model form uncertainty in Reynolds-averaged turbulence models with Bayesian deep neural networks
- Physics-informed covariance kernel for model-form uncertainty quantification with application to turbulent flows
- Subgrid-scale model for large-eddy simulation of isotropic turbulent flows using an artificial neural network
- Linear and nonlinear sensor placement strategies for mean-flow reconstruction via data assimilation
- Hierarchical deep-learning neural networks: finite elements and beyond
- Assimilation of disparate data for enhanced reconstruction of turbulent mean flows
- Determining \textit{a priori} a RANS model's applicable range via global epistemic uncertainty quantification
- Using machine learning to detect the turbulent region in flow past a circular cylinder
- Uncertainty quantification analysis for simulation of wakes in wind-farms using a stochastic RANS solver, compared with a deep learning approach
- Conditioning and accurate solutions of Reynolds average Navier–Stokes equations with data-driven turbulence closures
- Bayesian calibration of the constants of the \(k-\varepsilon\) turbulence model for a CFD model of street canyon flow
- Data fusion for uncertainty quantification with non-intrusive polynomial chaos
- DAFI: An Open-Source Framework for Ensemble-Based Data Assimilation and Field Inversion
- Machine learning for vortex induced vibration in turbulent flow
- Adaptive regularisation for ensemble Kalman inversion
- The image-based multiscale multigrid solver, preconditioner, and reduced order model
- Subgrid modelling for two-dimensional turbulence using neural networks
- Bayesian estimates of parameter variability in the \(k-\varepsilon\) turbulence model
- Predictive RANS simulations via Bayesian model-scenario averaging
- Bayesian model-scenario averaged predictions of compressor cascade flows under uncertain turbulence models
- Regularized ensemble Kalman methods for inverse problems
- A turbulent eddy-viscosity surrogate modeling framework for Reynolds-averaged Navier-Stokes simulations
- HiDeNN-TD: reduced-order hierarchical deep learning neural networks
- HiDeNN-FEM: a seamless machine learning approach to nonlinear finite element analysis
- Combining direct and indirect sparse data for learning generalizable turbulence models
- Ensemble Gradient for Learning Turbulence Models from Indirect Observations
- Assessments of epistemic uncertainty using Gaussian stochastic weight averaging for fluid-flow regression
- A confidence-based aerospace design approach robust to structural turbulence closure uncertainty
- Optimized parametric inference for the inner loop of the multigrid ensemble Kalman filter
- Evaluation of ensemble methods for quantifying uncertainties in steady-state CFD applications with small ensemble sizes
- A bi-fidelity ensemble Kalman method for PDE-constrained inverse problems in computational mechanics
- Sub-grid scale model classification and blending through deep learning
Uses Software
This page was built for publication: Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier-Stokes simulations: a data-driven, physics-informed Bayesian approach
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q525932)
