Using Infrared laser absorption to measure hydrocarbon concentration in a lean-burn, stratified-charge, spark-ignition engine

Abstract

The operating range of lean-burn spark-ignition (SI) engines is limited by the cycle-to-cycle variability of the fuel concentration at or near the spark plug at ignition timing. An experimental investigation was undertaken to measure the temporal and spatial distribution of hydrocarbon (HC) concentration in a spark-ignition engine, using the infrared (IR) laser absorption at 3392 nm. The purposes were to establish whether there is a correlation between time-resolved HC measurements for a range of global air-to-fuel (A/F) ratios (A/F = 15.5–23) with the strength of the firing stroke and to establish how this varies with fuel port-injection strategies against either open (injection timing 30° crank angle (CA) after intake top dead center) or closed-valves (injection timing 180°CA after intake top dead center), respectively resulting in stratified and near-homogeneous charge distributions. The results showed that IR line-of-sight (LOS) averaged A/F ratio measurements yielded a good agreement with the global A/F ratio readings obtained by the linear air-to-fuel (LAF) zirconia-based sensor. Furthermore, the cyclic variability of the measurements of the fuel concentration increased with increasing A/F ratio. At A/F = 23, closed-valve injection strategy resulted in small spatial stratification of the fuel charge with an ensemble-averaged correlation coefficient of fluctuations of the IR LOS A/F ratios with fluctuations of peak in-cylinder pressure of 0.37. For open-valve injection strategy, which resulted in axial fuel mixture charge-stratification but no radial charge-stratification, a maximum correlation coefficient of the IR LOS A/F ratios with fluctuations of peak pressure of 0.34 was measured (at the measurement location closest to the spark). This correlation was reduced to 0.17 at locations furthest from the spark plug. Finally, results showed that at A/F = 23, fuel stratification can be used to control lean-burn SI combustion, while at A/F = 15.5, no significant difference was observed between the two injection timings. The consequences of these findings on the fuel distribution in the pentroof and the control of the lean limits of SI engines are discussed.