Perspectives on predicting and controlling turbulent flows through deep learning

Abstract

The current revolution in the field of machine learning is leading to many interesting developments in a wide range of areas, including fluid mechanics. Fluid mechanics, and more concretely turbulence, is an ubiquitous problem in science and engineering. Being able to understand and predict the evolution of turbulent flows can have a critical impact on our possibilities to tackle a wide range of sustainability problems (including the current climate emergency) and industrial applications. Here, we review recent and emerging possibilities in the context of predictions, simulations, and control of fluid flows, focusing on wall-bounded turbulence. When it comes to flow control, we refer to the active manipulation of the fluid flow to improve the efficiency of processes such as reduced drag in vehicles, increased mixing in industrial processes, enhanced heat transfer in heat exchangers, and pollution reduction in urban environments. A number of important areas are benefiting from ML, and it is important to identify the synergies with the existing pillars of scientific discovery, i.e., theory, experiments, and simulations. Finally, I would like to encourage a balanced approach as a community in order to harness all the positive potential of these novel methods.