Exploratory physics-based turbulent flow modification using distributed computing architectures
Loading...
Authors
Karlik, Brooks
Issue Date
2023
Type
Thesis
Language
Keywords
computing , control , shock boundary layer interaction
Alternative Title
Abstract
Turbulent flows and their properties are of great interest in science and engineering. We propose and study the effects of flow modification strategies in three-dimensional incompressible and compressible turbulent flows. For incompressible flows, we present a novel physics-based control design for the selective modification of conserved flow quantities. In particular, we automatically identify forcing structures that selectively alter the energy and helicity of the flow, and the scales at which they are applied. We find that selective helicity modification may excite certain Kelvin (twist) modes within individual vortex tubes. Further, we present python bindings to a supersonic GPU accelerated turbulent shock boundary layer solver in an effort to leverage application of different feedback control techniques in future work. To orchestrate the execution of the hundreds of simulations required in this thesis, we present a distributed computing platform to allow for the scheduling of jobs across a cluster of computers without limitations on shared-memory or shared-filesystems.
Description
Citation
Publisher
License
Creative Commons Attribution-NonCommercial 4.0 United States