Multi-Task Learning based Convolutional Models with Curriculum Learning for the Anisotropic Reynolds Stress Tensor in Turbulent Duct Flow

TL;DR

This paper introduces a multi-task convolutional neural network with curriculum learning to improve the prediction of anisotropic Reynolds stress tensors in turbulent duct flows, advancing data-driven turbulence modeling.

Contribution

It presents a novel multi-task fully convolutional neural network architecture combined with curriculum learning for turbulence modeling.

Findings

Accurately predicts normalized anisotropic Reynolds stress tensor.

Demonstrates effectiveness of curriculum learning in turbulence modeling.

Enhances generalization over traditional models.

Abstract

The Reynolds-averaged Navier-Stokes (RANS) equations require accurate modeling of the anisotropic Reynolds stress tensor. Traditional closure models, while sophisticated, often only apply to restricted flow configurations. Researchers have started using machine learning approaches to tackle this problem by developing more general closure models informed by data. In this work we build upon recent convolutional neural network architectures used for turbulence modeling and propose a multi-task learning-based fully convolutional neural network that is able to accurately predict the normalized anisotropic Reynolds stress tensor for turbulent duct flows. Furthermore, we also explore the application of curriculum learning to data-driven turbulence modeling.