Turbulent shear flows are characterized by an interplay of many scales that describe persistent, quasi-invariant motion as well as violent, intermittent events. A data-driven computational framework, based on the decomposition of an embedded phase-space trajectory together with a community-identification step, will be introduced to properly describe and analyze these slow-fast dynamics. The framework combines elements of dynamic system theory with network analysis, and is applied to data-sequences from a reduced model of the turbulent self-sustaining process (SSP) in wall-bounded shear flows. Its effectiveness in detecting and quantifying structures and in laying the foundation for their targeted manipulation will be assessed.
Copyright. All Rights Reserved.