Despite significant progress in our understanding of turbulence over the last decades, predicting the small-scale features of turbulent flows using theoretical and computational tools remains a challenge. Small-scale turbulence plays a key role in phenomena such as rain formation, combustion, and mixing. A central issue is intermittency: the tendency of turbulent motion to be distributed unevenly in space and time, concentrated in patches of high activity surrounded by calmer flow. As a result, the statistics of small-scale turbulence are characterized by frequent extreme events and heavy-tailed distributions. In this talk, I present ensemble models to capture the intermittent features of small-scale turbulence. The ensemble approach mixes the statistics of turbulence simulations with varying parameters in order to reproduce the heavy-tailed distributions of flows of interest. In the first part, I consider turbulence simulations with slowly varying energy injection rates as an example of large-scale intermittency. The effect of these variations on small-scale statistics can be understood through different levels of ensemble modeling, shedding light on questions of turbulence universality. In the second part, I consider internal intermittency, i.e., the inhomogeneous redistribution of energy by the intrinsic flow dynamics, which intensifies with Reynolds number. I present an ensemble framework for predicting small-scale statistics of high-Reynolds-number turbulence using ensembles of smaller turbulence simulations. Overall, the ensemble approach provides a unifying perspective on turbulence intermittency.
Zoom Meeting: https://zoom.us/j/92718197851?pwd=Kq2sVih6BSfk2XvAV7WH9tTmEaoNTo.1
Meeting ID: 927 1819 7851
Passcode: 202030