Particulate Matter Emissions in Gasoline Direct-Injection Spark-Ignition Engines: Sources, Fuel Dependency, and Quantities

• Published in: Fuel (Guildford) 2023-04, Vol.338, p.127198, Article 127198
• Main Authors: Singh, Eshan, Kim, Namho, Vuilleumier, David, Skeen, Scott, Cenker, Emre, Sjöberg, Magnus, Pickett, Lyle
• Format: Article
• Language: English
• Subjects: Direct-injected spark ignition engine; Fuel effect; Optical diagnostics; Soot emissions; Spray
• ISSN: 0016-2361; 1873-7153
• DOI: 10.1016/j.fuel.2022.127198

•Experiments were conducted in metal and optical engine, and optically accessible spray chamber to investigate the dependence of engine-out soot on fuel composition.•Particulate matter index was found to not represent engine out soot emissions, especially at transient engine operation.•Fuels having similar chemical composition differed in their spray structure leading to different source of soot formation inside the engine.•Fouling of injectors significantly increased the engine-out soot emissions.•E30 combustion showed pool fires. The difference in time-scales of the observations suggest that pyrolysis pathways were not involved in pool fires. This study investigates sources of particulate matter (PM) from a direct-injected spark-ignition (DISI) engine, all of which are likely the result of poor mixture formation, although through varying mechanisms. Nine unique gasoline formulations, including both oxygenated and non-oxygenated fuels, with varying distillation curves and concentrations of different hydrocarbon classes were employed to investigate the effect of fuel on soot formation. The engine experiments were further supported by reacting and non-reacting experiments in a spray-chamber. The metal-engine experimental results indicated that engine-out PM emissions of the nine fuels generally conforms to their sooting tendency, here quantified by Particulate Matter Index (PMI). Discrepancies with the engine-out PM emissions and PMI were observed under transient operating condition, which is not reflective of the test conditions used to formulate the PMI. A diisobutylene surrogate blend and a high-olefin full-boiling range fuel were investigated further in the optically-accessible engine, showing soot generation from different sources. These differences came from the variation in their spray characteristics, as confirmed in the non-reactive spray-chamber experiments. While the earlier tests were conducted with a fouled injector, the injector was later cleaned and fuels were tested in the optical engine. High olefin fuel and diisobutylene blend showed significantly lower soot emissions with clean injector. E30 fuel exhibited spray impingement and subsequent diffusive combustion, known as pool fire, on the piston top characterized by a yellow flame due to soot luminosity. A final round of experiments in the spray-chamber mimic this pool fire mechanism. A stoichiometric flame was initiated after a spray has impinged on a wall and formed a film. Results showed