A series of experiments were conducted in an optical engine under lean natural gas spark ignited pre-chamber engine operation to study the end-gas autoignition process. Fundamental to these experiments was the ability to image autoignition events as fully as possible, allowing for correlation of optical imaging results with pressure data. To improve imaging of the autoignition process, a new two-hole pre-chamber nozzle was designed that centrally located end-gas within the piston window’s field of view while minimizing end-gas outside of the field of view. Using methane at non-autoigniting intake conditions of 100° C and 200 kPa, end-gas regions were observed to remain isolated within the field of view at +10 CAD. Using a lower methane number fuel, MN 65, operating conditions were identified that resulted in end-gas autoignition. It was demonstrated that end-gas autoignition could be reliably produced for each fired cycle within the field of view. Correlating the high-speed images with the measured pressure data, it was confirmed that the rapid increases in heat release rate of the cycle corresponded to the imaged end-gas autoignition event.
A larger set of operating conditions were identified that produced autoignition using MN 65 and MN 80 fuels with nominal peak fired in-cylinder pressures of 60, 80, and 100 bar, each run at equivalence ratios of 0.6, 0.7, and 0.8. For the 80 bar cases a thorough assessment of the autoignition behavior was conducted in two steps. First, a set of high-speed images were taken for each 80 bar case and analyzed by documenting the crank angle when autoignition appeared to start and end; general combustion and autoignition characteristics were also noted. Second, an independent pressure data analysis metric was developed that determined when the start and end of autoignition occurred; this metric was validated against the corresponding high-speed iii images. Using the pressure data analysis metric, the larger set of autoigniting 80-bar peak fired data was analyzed. It was observed that the earlier autoignition occurred, larger fractions of heat release occurred which produced both larger peak heat release rates and peak main chamber pressures. The duration of the autoignition event was found to be approximately constant between both fuels and the three equivalence ratios, indicating that earlier autoignition events consume fuel energy at higher rates. The crank angle corresponding to start of autoignition was used to determine the pressure and isentropic end-gas temperature at this time. It was found that the later autoignition correlated with the lower the pressure and end-gas temperature at the start of autoignition. Regardless of equivalence ratio, the variation in isentropic end-gas temperatures were shown to vary no more than 30° C for each fuel, with MN 65 and MN 80 having average temperatures of 1008 and 1048 K, respectively.