Monitoring and Diagnosis of Combustion Instability Using Laser-Induced Plasma in High-Pressure Environments

Abstract

With increasing concerns regarding global warming, the aeronautical sector has put prime importance on minimizing pollutant emissions. Operating the engines in a lean, high-pressure combustion regime is a favoured solution that has been at the centre of much research in past decades. However, lean flames become highly sensitive to coupling pressure fluctuations with the heat release rate along the flame front, also known as combustion instabilities. This feedback mechanism can be caused by many diverse factors that need to be controlled in order to operate combustion engines in safe regimes while optimizing gas emissions. More recently, hydrogen-fuelled combustion has also emerged as a potential long-term solution to reduce harmful gas release in the atmosphere. However, the altered flame compositions have led to significant dynamic changes, which are not fully understood to this day. The specific objectives of this thesis are to build reliable monitoring and diagnosis methods in highly pressurized environments using laser-induced plasmas.

The first part of the thesis focuses on presenting the combustor and the developed optical diagnostic setup. A full description of the modifications made to the supply line for the diagnosis technique is also presented. Typically, laser-induced breakdown suffers from instability in highly pressurized environments, generating a noisy and broadened spectrum. An improved plasma emission signal is obtained by reducing the plasma's temporal pulse width by limiting the photon absorption process. Four databases are constructed from the emission of the modulated nanoseconds laser-induced breakdown over a laminar flame with varied compositions; equivalence ratio and hydrogen content. The collected spectra are used for calibrating surrogate models after undergoing a Proper Orthogonal Decomposition and a sensitivity analysis. The models can predict three distinct flame properties when a single spectrum is given as an input; pressure, equivalence ratio and adiabatic temperature. It is proven that chopping the laser pulse width improves prediction performances for pressurized flames.

Finally, the last part of this thesis presents a novel combustion instability diagnosis technique. A laser-induced breakdown is used as a source of high-pressure disturbance that only lasts for a short period of time, with minimal time delays. Initially, cold flow tests are carried out before the reacting tests. Various flames composition are tested and their frequency response are first compared to the cold flow identification and then to the equivalent flame conditions with the plasma actuation inside the supply line. In that case, the dynamic responses are categorized with respect to the flame's mean noise level. Pressure recordings are analyzed in the frequency domain to facilitate the understanding of the added acoustic energy's impact on the system's non-reacting and reacting frequency response. A few conclusions are reached. Firstly, the plasma actuation intensifies the supply line frequencies and facilitates the identification of the system's instability frequencies, mostly dominated by the supply line geometry. Secondly, there is a constant frequency shift present between the reacting cases and the cold flow with the plasma actuation results. If the flame's frequency matches the system's, then there is a high risk that the combustion instability will be severe. Finally, the plasma's pressure perturbation can have a stabilizing effect on the flame's frequency when a particular temporal pattern is present. Those findings promise a practical application of laser-induced plasma to turbine engines as a reliable mean for an affordable and simple combustion instability diagnosis tool, and potentially a control tool.