The Computer as Turbulence Researcher (Prof. Javier Jiménez) – Part 1
This lecture was given by Prof. Javier Jiménez, Universidad Politecnica de Madrid, Spain in the framework of the von Karman Lecture Series on Machine Learning for Fluid Mechanics organized by the von Karman Institute and the Université libre de Bruxelles in February 2020. Using the identification of causally significant flow structures in two-dimensional turbulence as an example, this chapter explores how far the usual procedure of planning experiments to test hypotheses can be substituted by `blind' randomised trials, and notes that the increased efficiency of computers is beginning to make such a `Monte Carlo' approach practical in fluid mechanics. The processes of data generation, model creation, validation and verification are described in some detail. Although the purpose of the lecture is to explore the procedure, rather than to develop new models for two-dimensional turbulence, it is intriguing that the Monte Carlo process naturally led to the consideration of vortex dipoles as building blocks of the flow, on a par with the more classical individual vortex cores. Although not completely novel, this `spontaneous' discovery supports the claim that an important advantage of randomised experiments is to bypass researcher prejudice and paradigm lock. The method can be extended to three-dimensional flows in practical times.