Virtually all combustion chambers are subject to instabilities. Consequently there is a need to better understand them so as to control them. A possibility is to simulate the reactive flow within a combustor with the Large-Eddy Simulation (LES) method. However LES results come at a tremendous computational cost. Another route is to reduce the complexity of the problem to a simple thermoacoustic Helmholtz wave equation, which can be solved in the frequency domain as an eigenvalue problem. The coupling between the flame and the acoustics is then taken into account via proper models. The main drawback of this latter methodology is that it relies on the zero-Mach number assumption. Hence all phenomena inherent to mean flow effects are neglected. The present thesis aims to provide a novel strategy to introduce back some mean flow effects within the zero-Mach number framework. In a first part, the proper way to impose high-speed elements such as a turbine is investigated. The second part focuses on the coupling between acoustics and temperature heterogeneities that are naturally generated at the flame and convected downstream by the flow. Such phenomenon is important because it is responsible for indirect combustion noise that may drive a thermoacoustic instability. A Delayed Entropy Coupled Boundary Condition (DECBC) is then derived in order to account for this latter mechanism in the framework of a Helmholtz solver where the baseline flow is assumed at rest. In the last part, a realistic aero-engine combustor that features a mixed acoustic/entropy instability is studied. The methodology developed in the present thesis is tested and compared to LES computations. It is shown that computations with the Helmholtz solver can reproduce a complex combustion instability, and that this latter methodology is a potential tool to design new combustors so as to predict and avoid combustion instabilities.
展开▼
