Experimental metrics for identifying origins of combustion variability during spark-assisted compression ignition

Abstract Spark-assisted compression ignition, SACI, can be used to control the combustion phasing of compression-ignition gasoline engines. However, implementation of this technique can be confounded by cyclic variability. The purpose of this paper is to define experimental metrics that describe the SACI process and to demonstrate the use of these metrics for identifying the source(s) of cyclic variability during the SACI process. This study focused on a light load condition (7 mg/cycle, 200 kPa i.m.e.p.), where spray-guided direct fuel injection with spark ignition and an exhaust-rebreathing strategy was employed to achieve flame propagation, which led to compression ignition. This study employed a combination of measurements including pressure-based heat-release analysis, spark-discharge voltage/current measurements, and cycle-resolved combustion imaging. Based on these measurements, four distinct combustion periods were identified; namely, the spark discharge, the early kernel growth (EKG), flame propagation, and the compression ignition periods. Metrics were defined to characterize each period and used to identify the contribution of each period to the cyclic variability of the main heat release. For the light load condition studied here, the EKG period had the largest effect on the crank angle (CA) position of 50 per cent mass burned, CA50. The spark-discharge event may affect CA50 indirectly through its influence on EKG. However, this could not be definitively assessed here since the camera was incapable of recording both the spark-discharge event and the flame images during cycles of the same tests.