OATAO - Open Archive Toulouse Archive Ouverte Open Access Week

Low-Order Modeling and High-Fidelity Simulations for the Prediction of Combustion Instabilities in Liquid Rocket Engines and Gas Turbines

Laurent, Charlelie. Low-Order Modeling and High-Fidelity Simulations for the Prediction of Combustion Instabilities in Liquid Rocket Engines and Gas Turbines. PhD, Energétique et Transferts, Institut National Polytechnique de Toulouse, 2020

[img]
Preview
(Document in English)

PDF (Author's version) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
5MB

Abstract

Over the last decades, combustion instabilities have been a major concern for a number of industrial projects, especially in the design of Liquid Rocket Engines (LREs) and gas turbines. Mitigating their effects requires a solid scientific understanding of the intricate interplay between flame dynamics and acoustic waves that they involve. During this PhD work, several directions were explored to provide a better comprehension of flame dynamics in cryogenic rocket engines, as well as more efficient and robust numerical methods for the prediction of thermoacoustic instabilities in complex combustors. The first facet of this work consisted in the resolution of unstable thermoacoustic modes in complex multi-injectors combustors, a task that often requires a number of simplifications to be computationally affordable. These necessary physics-based assumptions led to the growing popularity of acoustic Low-Order Models (LOMs), among which Galerkin expansion LOMs have displayed a promising efficiency while retaining a satisfactory accuracy. Those are however limited to simple geometries that do not incorporate the complex features of industrial systems. A major part of this work therefore consisted first in clearly identifying the mathematical limitations of the classical Galerkin expansion, and then in designing a novel type of modal expansion, named a frame expansion, that does not suffer from the same restrictions. In particular, the frame expansion is able to accurately represent the acoustic velocity field, near non-rigid-wall boundaries of the combustor, a crucial ability that the Galerkin method lacks. In this work, the concept of surface modal expansion is also introduced to model topologically complex boundaries, such as multi-perforated liners encountered in gas turbines. These novel numerical methods were combined with the state-space formalism to build acoustic networks of complex systems. The resulting LOM framework was implemented in the code STORM (State-space Thermoacoustic low-ORder Model), which enables the low-order modeling of thermoacoustic instabilities in arbitrarily complex geometries. The second ingredient in the prediction of thermoacoustic instabilities is the flame dynamics modeling. This work dealt with this problem, in the specific case of a cryogenic coaxial jet-flame characteristic of a LRE. Flame dynamics driving phenomena were identified thanks to three-dimensional Large Eddy Simulations (LES) of the Mascotte experimental test rig where both reactants (CH4 and O2) are injected in transcritical conditions. A first simulation provides a detailed insight into the flame intrinsic dynamics. Several LES with harmonic modulation of the fuel inflow at various frequencies and amplitudes were performed in order to evaluate the flame response to acoustic oscillations and compute a Flame Transfer Function (FTF). The flame nonlinear response, including interactions between intrinsic and forced oscillations, were also investigated. Finally, the stabilization of this flame in the near-injector region, which is of primary importance on the overall flame dynamics, was investigated thanks to muulti-physics two-dimensional Direct Numerical Simulations (DNS), where a conjugate heat transfer problem is resolved at the injector lip

Item Type:PhD Thesis
Uncontrolled Keywords:
Institution:Université de Toulouse > Institut National Polytechnique de Toulouse - Toulouse INP (FRANCE)
Laboratory name:
Research Director:
Poinsot, Thierry and Gicquel, Laurent
Statistics:download
Deposited On:15 Feb 2021 14:28

Repository Staff Only: item control page