Search

This paper presents a theoretical and experimental study of direct fuel injection at conditions relevant to spark ignition (SI) and compression ignition (CI) engines. The focus of this work is to identify the conditions under which fuel droplet formation should occur or be suppressed. An experimental investigation of the injection of sub- and supercritical propane into gaseous nitrogen is first discussed. This includes study of one case in which the fuel remained supercritical with respect to temperature and pressure throughout the injection event, and which appears to be the first time that truly supercritical hydrocarbon fuel injection is examined experimentally. A nondimensional parameter τ representing the ratio of the time scales of droplet formation and droplet evaporation is also proposed and used to explain the observed occurrence or suppression of fuel droplets at different conditions. While instants of τ < 1 suggest that droplets should always be observed in any plausible SI or CI engine design, τ > 1 also occurs during the bulk delivery of heavier hydrocarbons in CI engines. In such cases, this should justify simplified modeling of the spray as a dense fluid that mixes with its surroundings, ignoring droplet transport during the less important parts of the injection event.

This paper presents an experimental and numerical investigation of the autoignition of ethanol blended with several primary reference fuels (PRFs) and toluene reference fuels (TRFs) in a Cooperative Fuel Research (CFR) engine. Autoigniting, in-cylinder pressure traces are acquired under standard research octane number (RON) conditions. Equivalent, non-autoignitiing traces are obtained by adding a small amount of tetraethyl lead (TEL) to each fuel to suppress knock, and these are used to calibrate a two-zone engine model of autoignition. The simulated autoignition timing of the PRFs without ethanol, TRFs without ethanol, and ethanol/PRF and ethanol/TRF blends are then compared to those measured in the engine. These results suggest that the incorporation of residual NO significantly improves the agreement between simulations and experiment for all cases with no or low ethanol content. Ethanol also appears to suppress the low-temperature activity of n-heptane. Both of these results are consistent with previous, more fundamental studies. However, close agreement between the simulated and measured autoignition timing across all fuel blends is not observed. Thus, the modeling does not comprehensively explain the significant synergism and antagonism observed in a recent study of the octane numbers of these fuels (Foong, T. M.; Morganti, K. J.; Brear, M. J.; da Silva, G.; Yang, Y.; Dryer, F. L.The octane numbers of ethanol blended with gasoline and its surrogates. Fuel2014, 115, 727−739, DOI: 10.1016/j.fuel.2013.07.105), demonstrating that further research is required. These results also contribute to a growing body of evidence suggesting the importance of NO chemistry, which should be included in kinetic simulations that attempt to model autoignition and knock in real engines.