Modelling the Pressure Strain Correlation in turbulent flows using Deep Neural Networks

TL;DR

This paper develops neural network models for the pressure strain correlation in turbulent flows, trained on DNS data, aiming to improve turbulence stress predictions over classical models.

Contribution

It introduces a physics-informed neural network approach for pressure strain correlation modeling at the Reynolds stress level, extending applications beyond eddy viscosity models.

Findings

Neural network models accurately predict pressure strain correlation.

Models generalize well to different flow conditions and Reynolds numbers.

Proposed framework shows potential for improved turbulence modeling.

Abstract

The pressure strain correlation plays a critical role in the Reynolds stress transport modelling. Accurate modelling of the pressure strain correlation leads to proper prediction of turbulence stresses and subsequently the other terms of engineering interest. However, classical pressure strain correlation models are often unreliable for complex turbulent flows. Machine learning based models have shown promise in turbulence modeling but their application has been largely restricted to eddy viscosity based models. In this article, we outline a rationale for the preferential application of machine learning and turbulence data to develop models at the level of Reynolds stress modeling. As the illustration We develop data driven models for the pressure strain correlation for turbulent channel flow using neural networks. The input features of the neural networks are chosen using physics based rationale. The networks are trained with the high resolution DNS data of turbulent channel flow at different friction Reynolds numbers. The testing of the models are performed for unknown flow statistics at other friction Reynolds numbers and also for turbulent plane Couette flows. Based on the results presented in this article, the proposed machine learning framework exhibits considerable promise and may be utilized for the development of accurate Reynolds stress models for flow prediction.