Deep Learning to advance the Eigenspace Perturbation Method for Turbulence Model Uncertainty Quantification

TL;DR

This paper presents a machine learning-enhanced approach to improve the Eigenspace Perturbation Method for quantifying turbulence model uncertainty in RANS simulations, demonstrating accurate identification of error-prone regions.

Contribution

It introduces a neural network-based method to predict discrepancies in Reynolds stress shapes, advancing uncertainty quantification in turbulence modeling.

Findings

Accurately identifies regions of modeling errors in turbulent flows.

Demonstrates improved uncertainty prediction over traditional methods.

Validates approach against DNS, LES, and experimental data.

Abstract

The Reynolds Averaged Navier Stokes (RANS) models are the most common form of model in turbulence simulations. They are used to calculate Reynolds stress tensor and give robust results for engineering flows. But RANS model predictions have large error and uncertainty. In past, there has been some work towards using data-driven methods to increase their accuracy. In this work we outline a machine learning approach to aid the use of the Eigenspace Perturbation Method to predict the uncertainty in the turbulence model prediction. We use a trained neural network to predict the discrepancy in the shape of the RANS predicted Reynolds stress ellipsoid. We apply the model to a number of turbulent flows and demonstrate how the approach correctly identifies the regions in which modeling errors occur when compared to direct numerical simulation (DNS), large eddy simulation (LES) or experimental results from previous works.