An 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. An ANN is used to establish a functional relation between the grid-scale flow field and the SGS stress without any assumption of the form of function. Data required for training and test of the ANN are provided by direct numerical simulation of a turbulent channel flow. It is shown that an ANN can establish a model similar to the gradient model. The correlation coefficients between the real SGS stress and the output of the ANN are comparable to or larger than similarity models, but smaller than a two-parameter dynamic mixed model. Large-eddy simulations using the trained ANN are also performed. Although ANN models show no advantage over the Smagorinsky model, the results confirm that the ANN is a promising tool for establishing a new subgrid model with further improvement.
ASJC Scopus subject areas
- Computational Mechanics
- Modelling and Simulation
- Fluid Flow and Transfer Processes