Formulating turbulence closures using sparse regression with embedded form invariance

Beetham, Sarah; Capecelatro, Jesse

doi:10.1103/physrevfluids.5.084611

Cited by 74 publications

(49 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In recent years, with the enhancement of computers' processing capability, the big data technology is developing rapidly and has been widely used in many branches of science and technology. Machine learning (ML) methods such as eXtreme gradient boosting (Xgboost), 38 support vector machine (SVM), 39 random forest, 40 and artificial neural network (ANN) 41‐46 have been increasingly used in the field of chemical engineering. It was around since the 1940s, 47 but it was not until the last two decades that the ANN was applied to engineering (e.g., the filtered drag development using ANN 32,41,46 ).…”

Section: Introductionmentioning

confidence: 99%

Data‐driven modeling of mesoscale solids stress closures for filtered two‐fluid model in gas–particle flows

2021

View full text Add to dashboard Cite

This study performs data‐driven modeling of mesoscale solids stress closures for filtered two‐fluid model (fTFM) in gas–particle flows via an artificial neural network (ANN) based machine learning method. The data used for developing the ANN‐based predictive data‐driven modeling framework is systematically filtered from fine‐grid simulations. The loss function optimization result reveals that coupling two loss functions promotes more accurate predictions of the mesoscale solids stresses than using a single loss function. Further comprehensive assessments of closure markers demonstrate a systematic dependence of the mesoscale solids stresses on the filtered particle velocity and its gradient as additional anisotropic markers, instead of using the conventional isotropic filtered rate of solid phase deformation as a closure marker. An optimal three‐marker mesoscale closure is thus proposed. Comparative analysis of the conventional filtered model and present three‐marker model shows that the data‐driven model can substantially enhance the prediction accuracy.

show abstract

Section: Introductionmentioning

confidence: 99%

Data‐driven modeling of mesoscale solids stress closures for filtered two‐fluid model in gas–particle flows

2021

View full text Add to dashboard Cite

show abstract

“…We apply a newly formulated modelling methodology, sparse regression with embedded form invariance (Beetham & Capecelatro 2020), to highly resolved Eulerian-Lagrangian data for fully developed CIT. The benefits of this methodology as compared with Neural Networks, which have become increasingly popular, are (i) interpretability of the resultant closures, since they are in a closed algebraic formulation, (ii) ease of dissemination to existing RANS solvers and (iii) robustness to very sparse training sets.…”

Section: Discussionmentioning

confidence: 99%

“…The six-term model, in turn, reduces overall model error by more accurately describing both components; however, this is most pronounced in the cross-stream direction (see figures 4(b) and 4(d)). In addition to discovering compact, algebraic models, sparse regression is also robust to sparse training data (Beetham & Capecelatro 2020). To illustrate this, a model was discovered using a sparse training dataset corresponding to (Ar, α p ) = [(1.8, 0.05), (5.4, 0.001), (18.0, 2.55)] and then tested using the remaining six cases.…”

Section: Drag Productionmentioning

confidence: 99%

“…In Schmelzer et al (2020) the models are written as tensor polynomials and built from a library of candidate functions. In Beetham & Capecelatro (2020) Galilean invariance of the resulting models are guaranteed through thoughtful tailoring of the feature space.…”

Section: Introductionmentioning

confidence: 99%

“…In this work, the sparse identification modelling framework of Beetham & Capecelatro (2020) is employed to develop multiphase closure models for homogeneous, gravity-driven gas-solid flows. Eulerian-Lagrangian simulations are performed across a range of Archimedes numbers and volume fractions to provide training data.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Sparse identification of multiphase turbulence closures for coupled fluid–particle flows

2021

Self Cite

View full text Add to dashboard Cite

show abstract

Conventional anddata‐drivenmodeling of filtered drag, heat transfer, and reaction rate ingas–particleflows

Zhu

Ouyang

et al. 2021

AIChE Journal

View full text Add to dashboard Cite

This study presents conventional and artificial neural network‐based data‐driven modeling (DDM) methods to model simultaneously the filtered mesoscale drag, heat transfer and reaction rate in gas–particle flows. The dataset used for developing the DDM is filtered from highly resolved simulations closed by our recently formulated microscopic drag and heat transfer coefficients (HTCs). Results reveal that the filtered drag correction is nearly independent of filter size when including the filtered gas phase pressure gradient. We further find that the filtered HTC correction critically depends on the added filtered temperature difference marker while the filtered reaction rate correction shows weak dependence on the additional markers. Moreover, compared with conventional correlations, DDM predictions agree better with filtered resolved data. Comparative analysis is also conducted between existing HTC corrections and our work. Finally, the applicability of conventional and data‐driven models coupled with coarse‐grid computational fluid dynamics simulations for pilot‐scale (reactive) gas–particle flows is validated comprehensively.

show abstract

Formulating turbulence closures using sparse regression with embedded form invariance

Cited by 74 publications

References 41 publications

Data‐driven modeling of mesoscale solids stress closures for filtered two‐fluid model in gas–particle flows

Data‐driven modeling of mesoscale solids stress closures for filtered two‐fluid model in gas–particle flows

Sparse identification of multiphase turbulence closures for coupled fluid–particle flows

Conventional anddata‐drivenmodeling of filtered drag, heat transfer, and reaction rate ingas–particleflows

Contact Info

Product

Resources

About