Jae Sung Park
Date of this Version
Turbulence is an emergent phenomenon found throughout nature and engineering, alike. It plays a vital role in the aquatic locomotion of organisms, scalar mixing, fluid transport, shipping and transportation, and even the flow of biological fluids in the human body. Therefore, it is of utmost importance in both a practical and engineering sense to better understand turbulence with the goal of better controlling it. This dissertation focuses broadly on better understanding the underlying mechanisms behind wall-bounded turbulent flows, with an emphasis on exploiting those mechanisms for turbulence flow control.
We developed a numerical simulation to study the effect of slip surfaces on the dynamics of transitional and turbulent flows. Slip surfaces were found to promote the return of a turbulent flow to the laminar state. They also impact the transition to and from turbulence depending upon flow structure. The simulation was extended to study composite drag reduction of slip surfaces and polymer additives. An additive effect was observed due to the distinct drag reduction mechanisms of each individual method.
Using simulations and experiments, intermittent dynamics of turbulent flows were investigated which manifest in the form of low-drag events: events described by low levels of skin friction and three-dimensionality. Because these events exhibit desirable traits, they are targets for flow control techniques, and their characterization will hopefully inform more efficient flow control methods.
The minimal flow unit (MFU) approach to simulating turbulent flows was first popularized by the seminal 1991 work of Jiménez and Moin. Since then, the technique has become a powerful tool in teasing out underlying mechanisms of turbulent flows due to its ability to resolve the many scales in turbulence. While the technique faithfully captures the dynamics of most flows, there are questions surrounding larger Reynolds numbers. We investigate the efficacy of MFUs in promoting "healthy" turbulence and show that additional criteria should be put in place when simulating higher Reynolds number flows with MFUs.
Adviser: Jae Sung Park