Data-driven modelling of the Reynolds stress tensor using random forests with invariance

Kaandorp, Mikael; Dwight, Richard P.

doi:10.1016/j.compfluid.2020.104497

Cited by 82 publications

(67 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A representative subset of that is motivated by the reasons described above, namely the use of machine learning (ML) techniques that employ as the learning environment a high fidelity approach and as the injected environment a low fidelity one. In this regard, important contributions with respect to predictive turbulence modelling were developed in recent years by Duraisamy and co-workers (Tracey, Duraisamy & Alonso 2013Parish & Duraisamy 2016), by Ling and co-workers (Ling, Jones & Templeton 2016a;Ling, Kurzawski & Templeton 2016b;Ling et al 2016c), by Xiao and co-workers Wu et al 2017;Wu, Xiao & Paterson 2018), by Zhao et al (2020) and by Kaandorp & Dwight (2020). For recent reviews, the reader is referred to the works of Duraisamy, Iaccarino & Xiao (2019) and Brunton, Noack & Koumoutsakos (2020).…”

Section: Low Fidelity and High Fidelity Approaches To Turbulencementioning

confidence: 99%

See 1 more Smart Citation

Conditioning and accurate solutions of Reynolds average Navier–Stokes equations with data-driven turbulence closures

et al. 2021

View full text Add to dashboard Cite

show abstract

Section: Low Fidelity and High Fidelity Approaches To Turbulencementioning

confidence: 99%

“…2017; Wu, Xiao & Paterson 2018), by Zhao et al. (2020) and by Kaandorp & Dwight (2020). For recent reviews, the reader is referred to the works of Duraisamy, Iaccarino & Xiao (2019) and Brunton, Noack & Koumoutsakos (2020).…”

Section: Introductionmentioning

confidence: 97%

Conditioning and accurate solutions of Reynolds average Navier–Stokes equations with data-driven turbulence closures

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Here, we chose the popular ADAM optimizer (Kingma and Ba, 2014) with a relatively low value for the learning rate η (0.0001) and a relatively large batch size of 1000. As our training data contain a high amount of noise inherent to turbulence, these parameter choices were in our case needed to stabilize the training results and achieve good convergence.…”

Section: Ann Trainingmentioning

confidence: 99%

Development of a large-eddy simulation subgrid model based on artificial neural networks: a case study of turbulent channel flow

et al. 2021

View full text Add to dashboard Cite

Abstract. Atmospheric boundary layers and other wall-bounded flows are often simulated with the large-eddy simulation (LES) technique, which relies on subgrid-scale (SGS) models to parameterize the smallest scales. These SGS models often make strong simplifying assumptions. Also, they tend to interact with the discretization errors introduced by the popular LES approach where a staggered finite-volume grid acts as an implicit filter. We therefore developed an alternative LES SGS model based on artificial neural networks (ANNs) for the computational fluid dynamics MicroHH code (v2.0). We used a turbulent channel flow (with friction Reynolds number Reτ=590) as a test case. The developed SGS model has been designed to compensate for both the unresolved physics and instantaneous spatial discretization errors introduced by the staggered finite-volume grid. We trained the ANNs based on instantaneous flow fields from a direct numerical simulation (DNS) of the selected channel flow. In general, we found excellent agreement between the ANN-predicted SGS fluxes and the SGS fluxes derived from DNS for flow fields not used during training. In addition, we demonstrate that our ANN SGS model generalizes well towards other coarse horizontal resolutions, especially when these resolutions are located within the range of the training data. This shows that ANNs have potential to construct highly accurate SGS models that compensate for spatial discretization errors. We do highlight and discuss one important challenge still remaining before this potential can be successfully leveraged in actual LES simulations: we observed an artificial buildup of turbulence kinetic energy when we directly incorporated our ANN SGS model into a LES simulation of the selected channel flow, eventually resulting in numeric instability. We hypothesize that error accumulation and aliasing errors are both important contributors to the observed instability. We finally make several suggestions for future research that may alleviate the observed instability.

show abstract

“… 7 developed a random-forests-based model, which directly predicts the Reynolds stress anisotropy. Kaandorp 8 and Kaandorp and Dwight 9 proposed a tensor basis random forest (TBRF) model, which is the random forests analogue to the TBNN proposed by Ling et al . 6 .…”

Section: Background and Summarymentioning

confidence: 99%

A curated dataset for data-driven turbulence modelling

2021

View full text Add to dashboard Cite

The recent surge in machine learning augmented turbulence modelling is a promising approach for addressing the limitations of Reynolds-averaged Navier-Stokes (RANS) models. This work presents the development of the first open-source dataset, curated and structured for immediate use in machine learning augmented corrective turbulence closure modelling. The dataset features a variety of RANS simulations with matching direct numerical simulation (DNS) and large-eddy simulation (LES) data. Four turbulence models are selected to form the initial dataset: k-ε, k-ε-ϕt-f, k-ω, and k-ω SST. The dataset consists of 29 cases per turbulence model, for several parametrically sweeping reference DNS/LES cases: periodic hills, square duct, parametric bumps, converging-diverging channel, and a curved backward-facing step. At each of the 895,640 points, various RANS features with DNS/LES labels are available. The feature set includes quantities used in current state-of-the-art models, and additional fields which enable the generation of new feature sets. The dataset reduces effort required to train, test, and benchmark new corrective RANS models. The dataset is available at 10.34740/kaggle/dsv/2637500.

show abstract

Data-driven modelling of the Reynolds stress tensor using random forests with invariance

Cited by 82 publications

References 28 publications

Conditioning and accurate solutions of Reynolds average Navier–Stokes equations with data-driven turbulence closures

Conditioning and accurate solutions of Reynolds average Navier–Stokes equations with data-driven turbulence closures

Development of a large-eddy simulation subgrid model based on artificial neural networks: a case study of turbulent channel flow

A curated dataset for data-driven turbulence modelling

Contact Info

Product

Resources

About