Predicting Nanoparticle Formation in Aero-Engines Using Large Eddy Simulation: Model Development, Validation, and Application

To minimize the environmental impact of aviation and comply with legal regulations, reducing pollutant emissions is a critical challenge in the development of modern propulsion systems. Since the application of carbon-free alternatives for aviation is currently limited, carbon-based fuels, particularly sustainable aviation fuels (SAFs), are expected to remain the primary energy source for aviation. Since these fuels still produce CO₂ and soot emissions, accurate predictions of pollutant formation are essential already during the development phase of future aero-engines. This dissertation investigates the modeling of soot formation and evolution under realistic, engine-relevant conditions. The foundation of this work is the S-EQMOM (split-based extended quadrature method of moments) soot model, whose applicability for soot prediction in aero-engines is assessed using increasingly complex configurations. First, a laminar counterflow configuration is employed to investigate the fundamental processes of soot oxidation under varying mixing rates and their relation to soot breakthrough into the fuel-lean region. The model is then applied to a single-sector combustion chamber equipped with a real Rolls-Royce fuel injector. This lab-scale burner serves as basis for two studies, with simulation results validated against experimental data. The first study investigates the influence of fuel spray modeling on fuel-air mixture formation and, consequently, on soot formation. While global flow characteristics remain nearly unchanged, the accuracy of soot predictions increases significantly when more detailed information on the fuel spray is incorporated into the simulation. This highlights the importance of capturing the entire process chain, from primary breakup to soot formation. The second study focuses on soot formation and evolution throughout the entire combustor and compares the results with those of previous studies. A classification of the combustor into different zones enables an efficient analysis of the various soot pathways within the combustor. The results align well with those from the simplified counterflow setup and confirm the applicability of the S-EQMOM model to capture key flow phenomena. In the final study, soot emissions of a real BR710 aero-engine combustor are successfully simulated and validated against the certification data across the entire operating range of the engine. In summary, this dissertation presents a simulation framework for reliably predicting soot emissions from aero-engine combustors. It represents an important advancement towards predictive soot modeling and contributes significantly to reducing pollutant emissions in next-generation sustainable aero-engines.