Development of Flow Analysis Tools using Information Theory to Understand Flow Instability
Abstract
APPROVED FOR PUBLIC RELEASE Project Summary: Flow instability drives many of the operability issues found in Navy aviation technologies. For example, feedback between vortex shedding and shock-cell structure insupersonic jets drives screech. Additionally, coupling between vortex shedding and combustoracoustic modes drives thermoacoustic oscillations. Prediction of these instabilities can be difficultin flows with any level of complexity. Hydrodynamic instability theory, which can identify the unstable modes of a system and their modal characteristics, can be applied to a variety of flows.This theory allows us to predict mode frequencies, growth rates, and structural sensitivity, and provides mathematical insight into the instability modes and their mechanisms.However, flow complexity may make use of this theory intractable. For example, flows withfluctuating density gradients like partially-premixed flames or highly three-dimensional flows arenot well posed for the use of hydrodynamic instability analysis. Additionally, the data from flowsthatmay be well-suited for instability analysis is often incomplete or highly noisy due tolimitations in optical diagnostics. In this work, we proposed to use methods from informationtheory to investigate the instability behaviors of complex flows. Measures of information contentin signals from these flows such as mutual information and transfer entropy will be used toconstruct a general, multi-layer complex network representation of flow dynamics. This networkrepresentation will be interrogated to identify critical flow regions and other characteristicscontrolling flow dynamics.The proposed work includes four tasks. First, a suite of tools combining ideas from informationtheory and complex networks will be developed to understand flow instabilities and the spatiotemporal structure ofcritical regions that can influence their dynamics. Second, these tools will betested on systems of two-dimensional multi-element flow fields, which are an ideal case study forthese types of methods. By varying a few simple parameters in multi-element flows, includingtheir number and spacing, we can create a wide variety of high dimensional flow behaviors thatare still tractable to analyze using traditional stability analysis. Third, we will compare resultsobtained using stability analysis and the new information-based tools for these tractable flows toestablish links between the results of the two families of methods. Finally, the new methods willbe tested on more complex flows of Navy relevance, including swirling flows and twin jets toexplain dynamics.The final outcomes of the work are twofold. First, this work will implement novel analysismethodologies combining complex networks and information theory for hydrodynamic instabilityin flows that can cause operability issues in Navy-relevant systems. These methods will be usedto identify critical flow regions driving instabilities from time-resolved measurements and/orcomputations. This new capability will allow for fundamental understanding of drivingmechanisms, their dependence on key operational parameters, and identification of interventionsthat can influence flow dynamics in critical regions and thus enable suppression of these dynamics.These methods will be applicable to all data, including highly complex flows and low-quality ornoisy experimental data, which will provide insight into a range of key flow configurations. Thefact that this tool development is done in conjunction with hydrodynamic stability analysis fortractable configurations will provide a rigorous theoretical basis for interpretation of the resultsand extension to complex cases where stabilityanalysis is challenging. Second, these methods willbe used to understand behaviors of several key families of flows, including multi-element flows,swirling flows, and twin jets, enhancing our fundamental understanding of canonical flows ofNavy interest.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 09, 2024
- Source ID
- N000142412520
Entities
People
- Jacqueline O Connor
Organizations
- Office of Naval Research
- Pennsylvania State University
- United States Navy