Multi-Task Learning based Convolutional Models with Curriculum Learning for the Anisotropic Reynolds Stress Tensor in Turbulent Duct Flow
This work addresses the need for more general closure models in turbulence modeling for fluid dynamics applications, representing an incremental improvement over existing machine learning approaches.
The authors tackled the problem of accurately modeling the anisotropic Reynolds stress tensor in turbulent duct flows using machine learning, proposing a multi-task learning-based fully convolutional neural network that achieved accurate predictions, with curriculum learning explored to enhance data-driven turbulence modeling.
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.