Experimental and 0D Numerical Investigation of Ultra-Lean Combustion Concept to Improve the Efficiency of SI Engine

Abstract

International audience ; Recently, the car manufacturers are moving towards innovative Spark Ignition (SI) engine architectures with unconventional combustion concepts, aiming to comply with the stringent regulation imposed by EU and other legislators. The introduction of burdensome cycles for vehicle homologation, indeed, requires an engine characterized by a high efficiency in the most of its operating conditions, for which a conventional SI engine results to be ineffective. Combustion systems which work with very lean air/fuel mixture have demonstrated to be a promising solution to this concern. Higher specific heat ratio, minor heat losses and increased knock resistance indeed allow improving fuel consumption. Additionally, the lower combustion temperatures enable to reduce NOX production.Since conventional SI engines can work with a limited amount of excess air, alternative solutions are being developed to overcome this constraint and reach the above benefit. Among all these solutions, replacing the spark-plug with a Pre-Chamber (PC) ignition system is gaining increasing interest. For this architecture, the combustion process starts in the PC and propagates in the main-chamber in the form of multiple turbulent jets of hot gas, with high-turbulence level. This ensures stable flame propagation even under extremely lean mixtures.In this research activity, an ultra-lean PC SI engine is numerically and experimentally investigated to assess the potential improvement of the thermal efficiency for ultra-lean operations. To this aim, a research single cylinder engine, fuelled with gasoline, is tested at fixed load and speed, realizing an air / fuel ratio sweep. A 1D/0D model of the examined engine is implemented in a commercial modelling framework (GT-Power™), where "in-house developed" sub-models are embedded, simulating in-cylinder phenomena, such as combustion, turbulence, heat transfer and pollutant emissions.The numerical approach, preliminarily tuned against 3D simulations and experimental outcomes, ...