Effects of end-gas reactivity on fundamental combustion properties of primary reference fuel blends with ethanol at engine-relevant conditions

Abstract

Accurate fundamental combustion data under engine-relevant conditions for liquid fuels are essential for developing predictive models for large-scale simulations of practical combustion devices. However, in existing literature, such data are non-existent or scarce, due to various experimental complications that the high-pressure and temperature conditions introduce. To overcome these limitations, the present study involved a combined experimental and modeling effort to identify thermodynamic conditions that result in reliable and accurate data for flame propagation and autoignition. The study was carried out using the confined spherically expanding flame method for primary reference fuels and their blends with ethanol, which are relevant to gasoline formulation, and which exhibit low-temperature chemistry. Conditions that result in flames that are free of hydrodynamic instabilities were identified through multi-dimensional direct numerical simulations and were implemented in all experimental measurements. Ignition delay times were measured and modelled, and the effect of ethanol addition was quantified with excellent experimental repeatability and uncertainty. Various kinetic models were used, and the discrepancy between predicted and measured ignition delay times was found to increase with ethanol addition. While end-gas reactivity is essential for autoignition studies, it is undesirable when laminar flame speeds are measured. Such effects can be present at engine-relevant conditions and for the gasoline-relevant fuels considered herein. While it has been shown in modeling studies, the increase of the burning rate due to end-gas reactivity was quantified for the first time experimentally in the present study for selected conditions, emphasizing that proper vetting of laminar flame speed data under high-pressure and temperature conditions is required. Additionally, in modeling spark ignition engines, both low and high temperature kinetic models must be used, not only to capture autoignition but also to properly predict the burning rate.