Computational Modeling of Non-idealities in Gaseous and Multiphase Detonating Flows

Abstract

To enhance the efficiency and capability of aeropropulsion platforms, step changes in thermal efficiency and fuel burn, particulate emissions, and device thrust-to-weight ratio are of paramount importance. Such fundamental efficiency improvements may be attained through detonation-based combustion, in contrast to deflagration-based combustion traditionally found in gas turbines, rockets, and scramjets today. The rotating detonation engine (RDE) is one realization of detonation-based combustion, where a propagating detonation wave confined within a combustion chamber produces shock-based compression with rapid chemical heat release. The complex flow field within an RDE is fraught with non-idealities, ranging from incomplete propellant mixing and stratification, unsteady injector dynamics, secondary combustion, and multiple detonation waves. The goal of this dissertation was to 1) develop an understanding of the non-idealities within RDEs and their effects on device operation, 2) contribute high-fidelity data on detonation physics and RDE operation to the community, and 3) establish a framework for the assimilation of data from numerical and experimental campaigns and reduced-physics models for RDE performance estimation.

The numerical simulations were used to provide unprecedented insight into the physics of gaseous and multiphase detonating flows in RDE combustors. Canonical detonating flows revealed that discrete injection and mixture preburning result in detonation waves with diminished strengths, slower propagation speeds, and broader spatial structure than the theoretical expectation. The simulations of full-scale gas- and liquid-fueled RDEs performed here represent some of the highest fidelity computational studies of these systems, with advancements in the modeled degrees of freedom through increased cell resolutions, larger computational domains, high-order detailed chemical kinetics of hydrocarbon mixtures, and extended simulation times. These studies revealed that parasitic combustion ahead of the wave results in commensal combustion behind the wave; a notable portion of the heat released within an RDE is through slow and distributed deflagrative combustion processes (as much as 60 to 70 percent) rather than from detonation. However, secondary combustion, while decreasing the achievable thermal efficiency, may improve operability and the overall combustion efficiency due to the complex interaction of unsteady turbulent mixing and detonation wave propagation. Furthermore, multiple competing detonation and secondary waves influence the flow structure by altering local heat release and fuel/oxidizer mixing and are integral to RDE operation.

While small-scale processes are important for operational stability and dictate how heat is released within the system, variations at the small-scale do not produce propulsive performance changes of the same order. The operating mode cannot be determined a priori as it is a manifestation of the equilibrium between small- and large-scale processes. Due to this challenge, a modeling framework to construct a surrogate model for RDE performance using limited and disparate datasets was proposed and demonstrated on a characteristic detonation combustor. In essence, the modeling approach establishes a pathway to unravel the connection between RDE operating conditions and the propulsive performance of the device: a high-dimensional coupling. The multi-fidelity fit of a performance quantity of interest, such as specific thrust, yielded a solution with a tighter confidence interval than a standard Gaussian process fit using the information contained within data at different fidelity levels, such as experiments, numerical simulations, and reduced-physics models. Further, additional experiments and simulations can be commissioned to reduce model uncertainty in an iterative design loop. The outcomes from the studies within this dissertation help shape the community's current understanding of reacting flows within detonation engines.