Long-term predictions of turbulence by implicit U-Net enhanced Fourier neural operator

Abstract

Long-term predictions of nonlinear dynamics of three-dimensional (3D) turbulence are very challenging for machine learning approaches. In this paper, we propose an implicit U-Net enhanced Fourier neural operator (IU-FNO) for stable and efficient predictions on the long-term large-scale dynamics of turbulence. The IU-FNO model employs implicit recurrent Fourier layers for deeper network extension and incorporates the U-net network for the accurate prediction on small-scale flow structures. The model is systematically tested in large-eddy simulations of three types of 3D turbulence, including forced homogeneous isotropic turbulence, temporally evolving turbulent mixing layer, and decaying homogeneous isotropic turbulence. The numerical simulations demonstrate that the IU-FNO model is more accurate than other FNO-based models, including vanilla FNO, implicit FNO (IFNO), and U-Net enhanced FNO (U-FNO), and dynamic Smagorinsky model (DSM) in predicting a variety of statistics, including the velocity spectrum, probability density functions of vorticity and velocity increments, and instantaneous spatial structures of flow field. Moreover, IU-FNO improves long-term stable predictions, which has not been achieved by the previous versions of FNO. Moreover, the proposed model is much faster than traditional large-eddy simulation with the DSM model and can be well generalized to the situations of higher Taylor–Reynolds numbers and unseen flow regime of decaying turbulence.

Turbulence closure modeling with machine learning: a foundational physics perspective

“…Only the 'universal' and better-understood small scales are subject to closure modeling. There is significant literature on ML-LES approaches that aim to develop subgrid stress models from high-fidelity data [78][79][80][81][82][83][84][85][86][87][88][89][90][91][92][93][94]. This body of work can be broadly classified into parametric and non-parametric approaches.…”