177. TURBULENCE GENERATION AND PARTICLE DYNAMICS IN SHOCKED PARTICLE-LADEN FLOW

Department: Mechanical & Aerospace Engineering
Faculty Advisor(s): Sutanu Sarkar
Award(s): Honorable Mention

Primary Student
Name: Sean Lin Sheng Davis
Email: sldavis@ucsd.edu
Phone: 781-248-2318
Grad Year: 2015

Abstract
The 2010 volcanic explosion in Iceland cost $1.7 billion to the airline industry alone. The resultant ash cloud, spread by a variety of mechanisms, could wreak havoc on internal engine components. My research focuses on the elucidation of the fundamental mechanisms that are responsible for particle mixing and turbulence generation under these conditions. In particular, we aim to investigate instabilities in the interaction between an initially stationary cloud of particles and a moving shock. We hypothesize that the instabilities arise from a combination of shock-particle, particle-turbulence, shock-turbulence, and particle-particle interactions. These instabilities are strongly correlated with the large-scale coherent structures, mixing and transport processes in the fluid and particle phase. In a collaborative effort with scientists at the Russian Academy of Sciences, we are investigating shock-particle-turbulence interactions through high-fidelity computations, model development and experimental validation. In combination with an intricate physical composition, particle-laden turbulent flows with shocks exhibit a large range of time and spatial scales. These complexities pose high demands on both computations and experiments. Because of the high velocities, experimental observations are restricted to only large-scale effects such as particle dispersion and shock location. A computational approach provides significantly more detail, but model and numerical inaccuracies may limit the fidelity of the results. Particle-laden flows are commonly computed with Eulerian-Lagrangian (EL) methods and models, where the flow dynamics are determined in Eulerian frame, while particles are traced in their Lagrangian frame. Typically, these EL methods only compute flows using simplified models, simple grids and low order of accuracy techniques. These relatively low fidelity results are now showing bottlenecks in the computation and analysis of shocked turbulent particle-laden flows. Our high-fidelity, WENO based EL model was shown to capture shocks sharply, while accurately resolving small-scale shock-turbulence-particle interaction. Computations with the high-fidelity EL method on the interaction between a shock and a cloud of bronze particles were shown to be in good agreement with experiments.

« Back to Posters or Search Results