A Lagrangian view of turbulent mixing in channel flow through complex networks

Turbulent mixing and diffusion are investigated in the Lagrangian framework using tools proper to network theory. Starting from a database of trajectories of non-inertial particles obtained through direct numerical simulation of a channel flow, two innovative analysis tools are developed and discussed. The first tool aims to characterize the tendency of tracers to group in coherent structures, or clusters, despite being driven further and further away by the diffusive action of turbulence. The second tool deals with the diffusion and transport of particles in the only inhomogeneous direction of the channel flow, which is the wall normal one. In both tools particles (or sets of them) are represented as graph nodes, while different definition of links are given. Network theory offers a powerful framework to analyse complex dynamical behaviour in large datasets.