Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier–Stokes simulations: A data-driven, physics-informed Bayesian approach

Xiao, Heng; Wu, Jinlong; Wang, J.-X.; Sun, Rui; Roy, Christopher J.

doi:10.1016/j.jcp.2016.07.038

Cited by 270 publications

(243 citation statements)

References 51 publications

Supporting

Mentioning

221

Contrasting

Order By: Relevance

“…Machine learning has been used to identify and model discrepancies in the Reynolds stress tensor between a RANS model and high-fidelity simulations (Ling & Templeton 2015;Parish & Duraisamy 2016;Ling et al 2016b;Xiao et al 2016;Singh et al 2017;. Ling & Templeton (2015) compare support vector machines, Adaboost decision trees, and random forests to classify and predict regions of high uncertainty in the Reynolds stress tensor.…”

Section: Parsimonious Nonlinear Modelsmentioning

confidence: 99%

“…use random forests to built a supervised model for the discrepancy in the Reynolds stress tensor. Xiao et al (2016) leveraged sparse online velocity measurements in a Bayesian framework to infer these discrepancies. In related work, Parish & Duraisamy (2016) develop the field inversion and machine learning modeling framework, that builds corrective models based on inverse modeling.…”

Section: Parsimonious Nonlinear Modelsmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning for Fluid Mechanics

Brunton

Noack

Koumoutsakos

2020

Annu. Rev. Fluid Mech.

1,831

742

View full text Add to dashboard Cite

The field of fluid mechanics is rapidly advancing, driven by unprecedented volumes of data from experiments, field measurements, and large-scale simulations at multiple spatiotemporal scales. Machine learning presents us with a wealth of techniques to extract information from data that can be translated into knowledge about the underlying fluid mechanics. Moreover, machine learning algorithms can augment domain knowledge and automate tasks related to flow control and optimization. This article presents an overview of past history, current developments, and emerging opportunities of machine learning for fluid mechanics. We outline fundamental machine learning methodologies and discuss their uses for understanding, modeling, optimizing, and controlling fluid flows. The strengths and limitations of these methods are addressed from the perspective of scientific inquiry that links data with modeling, experiments, and simulations. Machine learning provides a powerful information processing framework that can augment, and possibly even transform, current lines of fluid mechanics research and industrial applications.

show abstract

Section: Parsimonious Nonlinear Modelsmentioning

confidence: 99%

Section: Parsimonious Nonlinear Modelsmentioning

confidence: 99%

Machine Learning for Fluid Mechanics

Brunton

Noack

Koumoutsakos

2020

Annu. Rev. Fluid Mech.

1,831

742

View full text Add to dashboard Cite

show abstract

“…These operations are based on the assumption that the averaging performed on it for any turbulent flow yield the magnitude of invariables during repeated averaging [16].…”

Section: Literature Review and Problem Statementmentioning

confidence: 99%

Formalization of design for physical model of the azimuth thruster with two degrees of freedom by computational fluid dynamics methods

Budashko

2017

EEJET

View full text Add to dashboard Cite

“…The following equations are the turbulent flow equations for incompressible flows. Continuity for the mean flow [10][11][12]:…”

Section: Theoretical Modelmentioning

confidence: 99%

“…The time-average transport equation for scalar ϕ is given by [10][11][12] ( ) ( ) To be capable of calculating turbulent flows with the RANS equations, it is essential to establish turbulence models to estimate the Reynolds stresses and the scalar transport terms and close the system of mean flow equations (1), (2-4), and (7). As the flow of high speed jet is highly turbulent, the high Reynolds number form of the k ε − turbulence model (i.e., RNG k ε − turbulence model) is utilized for the computational purpose [1].…”

Section: Theoretical Modelmentioning

confidence: 99%

Computational Fluid Dynamics Simulation of High Speed Jet Under Different Input Pressures

Gong¹

2017

IJHEP

View full text Add to dashboard Cite

Abstract:The aim of this study is to execute the computational fluid dynamics (CFD) simulation of high speed jet under different input pressures (i.e., 80, 120, 160, 200, and 240 MPa). In particular, this study focuses on the pressure distributions and streamlines of the orifice in high speed jet, primarily because the orifice plays a role in accelerating the flow of liquid, having significant effects on the working performance of high speed jet. Firstly, the two-dimensional geometric model of high speed jet is established on the basis of the actual operational conditions. Next, the unstructured grids of high speed jet are generated by means of ICEM CFD 16.0. Virtually, the computational fluid dynamics simulation of high speed jet is a two-phase flow (gas-liquid) problem, so the homogeneous (Eulerian-Eulerian) two-phase model is employed to carry out the gas-liquid interaction. Particularly, the turbulent flow computation of high speed jet is carried out with procedures based on the Reynolds-averaged Navier-Stokes (RANS) equations. As the flow of high speed jet is highly turbulent, the RNG k-ɛ turbulence model derived by Yakhot et al. (1992) is utilized in this study. Finally, the computational fluid dynamics (CFD) simulation of high speed jet is implemented by using the CFX-Solver in ANSYS CFX 16.0. The simulation results show that when liquid flows through the orifice, the pressure of flows decreases swiftly, whereas the velocity of flows skyrockets to the maximum value and then decreases slightly. In addition, the relationship between the working pressure and input pressure and the relationship between the working velocity and input pressure are achieved, which could provide certain theoretical guidance for predicting the working pressure and velocity of high speed jet based on real input pressures.

show abstract

Quantifying and reducing model-form uncertainties in Reynolds-averaged Navier–Stokes simulations: A data-driven, physics-informed Bayesian approach

Cited by 270 publications

References 51 publications

Machine Learning for Fluid Mechanics

Machine Learning for Fluid Mechanics

Formalization of design for physical model of the azimuth thruster with two degrees of freedom by computational fluid dynamics methods

Computational Fluid Dynamics Simulation of High Speed Jet Under Different Input Pressures

Contact Info

Product

Resources

About