Searching for turbulence models by artificial neural network

TL;DR

This paper explores using artificial neural networks to develop new subgrid-scale turbulence models in large-eddy simulations, trained on DNS data, showing promising correlation with actual SGS stresses.

Contribution

It introduces a neural network-based approach to model SGS stresses without assuming a specific functional form, trained on DNS data for turbulent channel flow.

Findings

ANN can produce SGS models similar to the gradient model.

Correlation coefficients are comparable to similarity models.

Performance is slightly below the two-parameter dynamic mixed model.

Abstract

Artificial neural network (ANN) is tested as a tool for finding a new subgrid model of the subgrid-scale (SGS) stress in large-eddy simulation. ANN is used to establish a functional relation between the grid-scale (GS) flow field and the SGS stress without any assumption of the form of function. Data required for training and test of ANN are provided by direct numerical simulation (DNS) of a turbulent channel flow. It is shown that ANN can establish a model similar to the gradient model. The correlation coefficients between the real SGS stress and the output of ANN are comparable to or larger than similarity models, but smaller than a two-parameter dynamic mixed model.