Search Results

This paper presents a numerical study of trace knocking combustion of ethanol/gasoline blends in a modern, single cylinder SI engine. Results are compared to experimental data from a prior, published work [1]. The engine is modeled using GT-Power and a two-zone combustion model containing detailed kinetic models. The two zone model uses a gasoline surrogate model [2] combined with a sub-model for nitric oxide (NO) [3] to simulate end-gas autoignition. Upstream, pre-vaporized fuel injection (UFI) and direct injection (DI) are modeled and compared to characterize ethanol's low autoignition reactivity and high charge cooling effects. Three ethanol/gasoline blends are studied: E0, E20, and E50. The modeled and experimental results demonstrate some systematic differences in the spark timing for trace knock across all three fuels, but the relative trends with engine load and ethanol content are consistent. Possible reasons causing the differences are discussed.

This paper examines the effect of charge cooling on the Research Octane Number (RON) of ethanol/gasoline blends. While gasoline is fully vaporized prior to entry into the engine in a standard RON test, significant charge cooling is observed for blends with high ethanol content, with the presence of a near-saturated and potentially two-phase air-fuel mixture during induction. Thus, the relative significance of the charge cooling and the autoignition chemistry cannot be determined from the standard RON test. In order to better delineate the effects of charge cooling and autoignition chemistry, a so-called ‘modified RON’ test is therefore devised in which the temperature of the air-fuel mixture entering the engine is fixed and representative of that observed for primary reference fuels (PRFs).