Study of premixed combustion induced by controlled hot surface ignition in stationary gas engines
- Scholl, Fino
- Blanca Giménez Olavarría Directora
- Maurice Kettner Codirector/a
Universidad de defensa: Universidad de Valladolid
Fecha de defensa: 04 de diciembre de 2017
- Francisco V. Tinaut Fluixá Presidente/a
- Magín Lapuerta Amigo Secretario/a
- Ulrich Spicher Vocal
Tipo: Tesis
Resumen
Lean-burn combustion of premixed air-natural-gas mixtures enables small-scale stationary gas engines to operate at high engine efficiency while emitting low levels of CO2 and pollutants. Increasingly stringent limits for nitrogen oxides, however, pose a major challenge to lean-burn operation if the aim is to meet the new limits through combustion-related measures without losses in engine efficiency. As alternative to spark ignition, a new Hot Surface Ignition (HSI) system was developed that facilitates a rapid and safe ignition of more diluted mixtures. The system essentially consists of a shielded ceramic glow plug, whose temperature can be controlled in a highly dynamic manner by adjusting the voltage applied, thereby allowing for the phasing of combustion to be adjusted during engine operation. The present work describes the systematic methodology adopted to study the process of HSI induced combustion. For an in-depth analysis of the ignition process and the subsequent flame propagation, a combined approach of 3D fluid dynamics and combustion simulation and engine experiments with multiple single-fibre optical accesses to the combustion chamber is pursued. Design of Experiments (DoE) methods are used to determine the ignition system’s operation limits and to study the effect of several operating parameters, like intake manifold pressure and relative air-fuel ratio, on the combustion process and the thermo-electric properties of the glow plug. The direct comparison to state-of-the-art prechamber spark ignition unfolds the positive and negative aspects of the current HSI design and allows for future steps in system develop-ment and optimisation to be derived.