Effects of Gasoline Composition and Octane Sensitivity on the Response of DISI Engine Knock to Variations of Fuel-Air Equivalence Ratio

Abstract

The need to avoid SI engine knock often comes with an efficiency penalty, motivating efforts to understand its causes. In this study, the relationship between knock and end-gas autoignition is examined based on experiments in a DISI engine with CR = 12, using three different gasoline fuels. The three gasolines are an alkylate blend (RON=98, MON=97), a blend with high aromatic content (RON=98, MON=88), and a blend of 30% ethanol by volume (RON=98, MON=87). Fuel/air-equivalence ratio (φ) sweeps show that the response of the knock limits to changes in φ varies between the fuels. Furthermore, it is observed that the statistics of knock and heat-release vary in a complex manner between the fuels.Since knock originates from end-gas autoignition, a robust heat-release-based metric of “trace autoignition” is developed to determine knock occurrence. The metric is based on the observation that autoigniting cycles exhibit a faster heat-release rate decay at the end of the combustion event, even if no knock oscillations are detected. In general, the acoustic knock trends match the trends of the “trace autoignition” metric. However, for rich operation, the Alkylate fuel needs to be operated with an average CA50 that is somewhat retarded relative to trace autoignition. Furthermore, the data reveal that the dependence of autoignition on φ varies in manner that relates to the RON-MON octane sensitivity of the fuel. For example, the Alkylate fuel with low octane sensitivity displays no benefit of fuel enrichment, which is in strong contrast to the two high octane-sensitive fuels. Further, the Alkylate fuel shows a strong reduction of its anti-knock quality for lean operation, which correlates with the development of low-temperature heat release.