Estimating uncertainty in homogeneous turbulence evolution due to coarse-graining

Mishra, Aashwin; Duraisamy, Karthik; Iaccarino, Gianluca

doi:10.1063/1.5080460

Cited by 26 publications

(11 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These factors make the k-ε model more appropriate for atmospheric simulations than the k-ω model. More limitations of k-ε and k-ω models are discussed in [31,32].…”

Section: Introductionmentioning

confidence: 99%

Modification of k-ε Model for Sensible Heat and Momentum Flux Reconstruction from Surface Temperature Data

Hrebtov

Bobrov

2022

Atmosphere

View full text Add to dashboard Cite

We present a modified k-ε model with a set of wall-functions suitable for reconstruction of sensible heat and momentum fluxes from the observations data (e.g., surface temperature evolution during the diurnal cycle). The modification takes into account stability and buoyancy effects in the Reynolds stress parametrization which affects turbulence production and turbulent heat flux. The single-cell and single-column versions of the model are presented. The model is tested based on CASES-99 observations data for dry ABL. It is shown that the presented modification improves the predictions of sensible heat flux magnitude and leads to a faster onset of a daytime instability, compared to the non-modified k-ε model and its scale-limited modification based on Monin-Obukhov similarity theory.

show abstract

“…These factors make the k-ε model more appropriate for atmospheric simulations than the k-ω model. More limitations of k-ε and k-ω models are discussed in [31,32].…”

Section: Introductionmentioning

confidence: 99%

Modification of k-ε Model for Sensible Heat and Momentum Flux Reconstruction from Surface Temperature Data

Hrebtov

Bobrov

2022

Atmosphere

View full text Add to dashboard Cite

show abstract

“…However any eddy viscosity based model, even with optimal coefficients inferred using machine learning, is incapable of capturing high degrees of turbulence anisotropy due to the nature of the linear eddy viscosity hypothesis inherent to the model. 39 This eddy viscosity hypothesis states that the turbulence anisotropy is a function of the instantaneous mean strain rate and thus must lie on the "plane strain" manifold of the barycentric triangle. 40 Consequently the anisotropy information from the high fidelity data is rendered ineffectual due to the form of the baseline model.…”

Section: Introductionmentioning

confidence: 99%

Modelling the Pressure Strain Correlation in turbulent flows using Deep Neural Networks

Panda¹,

Warrior²

2021

Preprint

View full text Add to dashboard Cite

The pressure strain correlation plays a critical role in the Reynolds stress transport modelling. Accurate modelling of the pressure strain correlation leads to proper prediction of turbulence stresses and subsequently the other terms of engineering interest. However, classical pressure strain correlation models are often unreliable for complex turbulent flows. Machine learning based models have shown promise in turbulence modeling but their application has been largely restricted to eddy viscosity based models. In this article, we outline a rationale for the preferential application of machine learning and turbulence data to develop models at the level of Reynolds stress modeling. As the illustration We develop data driven models for the pressure strain correlation for turbulent channel flow using neural networks. The input features of the neural networks are chosen using physics based rationale. The networks are trained with the high resolution DNS data of turbulent channel flow at different friction Reynolds numbers. The testing of the models are performed for unknown flow statistics at other friction Reynolds numbers and also for turbulent plane Couette flows. Based on the results presented in this article, the proposed machine learning framework exhibits considerable promise and may be utilized for the development of accurate Reynolds stress models for flow prediction.

show abstract

“…As an illustration DNS data is able to replicate the high degree of anisotropy in turbulent flows. However eddy viscosity based models are not capable of replicating a high degree of turbulence anisotropy due to the linear eddy viscosity hypothesis (Mishra, Duraisamy, and Iaccarino 2019).…”

Section: Introductionmentioning

confidence: 99%

Evaluation of machine learning algorithms for predictive Reynolds stress transport modeling

Panda¹,

Warrior²

2021

Preprint

View full text Add to dashboard Cite

The application machine learning (ML) algorithms to turbulence modeling has shown promise over the last few years, but their application has been restricted to eddy viscosity based closure approaches. In this article we discuss rationale for the application of machine learning with high-fidelity turbulence data to develop models at the level of Reynolds stress transport modeling. Based on these rationale we compare different machine learning algorithms to determine their efficacy and robustness at modeling the different transport processes in the Reynolds Stress Transport Equations. Those data driven algorithms include Random forests, gradient boosted trees and neural networks. The direct numerical simulation (DNS) data for flow in channels is used both as training and testing of the ML models. The optimal hyper-parameters of the ML algorithms are determined using Bayesian optimization. The efficacy of above mentioned algorithms is assessed in the modeling and prediction of the terms in the Reynolds stress transport equations. It was observed that all the three algorithms predict the turbulence parameters with acceptable level of accuracy. These ML models are then applied for prediction of the pressure strain correlation of flow cases that are different from the flows used for training, to assess their robustness and generalizability. This explores the assertion that ML based data driven turbulence models can overcome the modeling limitations associated with the traditional turbulence models and ML models trained with large amounts data with different classes of flows can predict flow field with reasonable accuracy for unknown flows with similar flow physics. In addition to this verification we carry out validation for the final ML models by assessing the importance of different input features for prediction.

show abstract

Estimating uncertainty in homogeneous turbulence evolution due to coarse-graining

Cited by 26 publications

References 28 publications

Modification of k-ε Model for Sensible Heat and Momentum Flux Reconstruction from Surface Temperature Data

Modification of k-ε Model for Sensible Heat and Momentum Flux Reconstruction from Surface Temperature Data

Modelling the Pressure Strain Correlation in turbulent flows using Deep Neural Networks

Evaluation of machine learning algorithms for predictive Reynolds stress transport modeling

Contact Info

Product

Resources

About