Data-Driven Modelling of the Reynolds Stress Tensor using Random Forests with Invariance

TL;DR

This paper introduces the Tensor Basis Random Forest (TBRF), a machine learning method that predicts Reynolds stress anisotropy in turbulence modeling, ensuring Galilean invariance and improving accuracy over previous approaches.

Contribution

The paper presents a novel, invariant machine learning algorithm for turbulence modeling that effectively predicts Reynolds stress anisotropy using a tensor basis within a random forest framework.

Findings

TBRF accurately predicts Reynolds stress anisotropy across various flows.

The method achieves good agreement with DNS/LES data for mean flow predictions.

TBRF outperforms previous neural network approaches in turbulence modeling.

Abstract

A novel machine learning algorithm is presented, serving as a data-driven turbulence modeling tool for Reynolds Averaged Navier-Stokes (RANS) simulations. This machine learning algorithm, called the Tensor Basis Random Forest (TBRF), is used to predict the Reynolds-stress anisotropy tensor, while guaranteeing Galilean invariance by making use of a tensor basis. By modifying a random forest algorithm to accept such a tensor basis, a robust, easy to implement, and easy to train algorithm is created. The algorithm is trained on several flow cases using DNS/LES data, and used to predict the Reynolds stress anisotropy tensor for new, unseen flows. The resulting predictions of turbulence anisotropy are used as a turbulence model within a custom RANS solver. Stabilization of this solver is necessary, and is achieved by a continuation method and a modified $k$-equation. Results are compared to the neural network approach of Ling et al. [J. Fluid Mech, 807(2016):155-166, (2016)]. Results show that the TBRF algorithm is able to accurately predict the anisotropy tensor for various flow cases, with realizable predictions close to the DNS/LES reference data. Corresponding mean flows for a square duct flow case and a backward facing step flow case show good agreement with DNS and experimental data-sets. Overall, these results are seen as a next step towards improved data-driven modelling of turbulence. This creates an opportunity to generate custom turbulence closures for specific classes of flows, limited only by the availability of LES/DNS data.