Increasing concern of the environmental impact of engine emissions coupled with a growing urgency to improve fuel efficiency has prompted significant efforts to develop high efficiency lean combustion strategies. Lean combustion strategies are those in which the engine is operated with excess air such that the air-to-fuel ratio (AFR) is greater than the stoichiometric condition. The extent to which excess air can be added in spark-ignition engines and still provide adequate combustion stability is limited by the ability of the ignition source to reliably ignite the mixture. One method which has demonstrated the capability of extending the lean operating limit while offering significant reductions in NOx emissions at ultra-lean operating conditions is pre-chamber initiated combustion.
In this work, optical measurements of the flame-shape during pre-chamber spark plug engine operation were conducted for a pre-chamber which was designed for use in a highly customized optical engine. The design of the pre-chamber was set with the intention of producing a similar pre-chamber torch jet to that of an MHI GS16R2-PTK natural gas fueled pre-chamber engine which has a bore diameter twice that of the optical engine used for these experiments. A commercially available spark plug was used in conjunction with a single fuel injector which supplied gaseous methane to both the pre- and main-chambers to initiate combustion in the pre-chamber. For the majority of the engine experiments, an overall equivalence ratio ranging from 0.64 to 0.59 was used with spark timings of -15 and -20 CAD.
Measurements of the shape of the pre-chamber torch jet and ensuing main-chamber flame front were performed using high-speed imaging of the broadband and OH chemiluminescence signals. An equivalence ratio and engine load sweep was performed in order to determine the appropriate operating conditions for subsequent OH planar laser induced fluorescence (PLIF) imaging experiments. High-speed images of broadband and OH chemiluminescence signals for the highest equivalence ratio tested (ϕ=0.64) showed that the pre-chamber jet produced a more volumetric interaction with the main-chamber and resulted in a greater fraction of the area burned in the main-chamber at times earlier than those observed for lower equivalence ratios. The high-speed images, heat release rate (HRR) results and plots of the fractional area burned showed minimal interaction of the pre-chamber jet with the main-chamber at the lowest equivalence ratio of 0.45 and also indicated that substantial main-chamber heat release was not observed until an equivalence ratio of 0.55.
Imaging of the pre- and main-chamber combustion events was also investigated using OH PLIF. Image timings of -12, +8 and +20 CAD were selected to capture the early pre-chamber jet, peak cylinder pressure, and late combustion events respectively. The images acquired at -12 and +8 CAD revealed aspects of the three-dimensional nature of the developing pre-chamber plume and main-chamber flame that were not apparent in the high-speed image sets. From images acquired at the latest image time of +20 CAD, it was also observed that portions of the cylinder remained unburnt for the operating conditions which had intake pressures of 200 and 250 kPa while unburnt regions at the +20 CAD image time were not apparent for lower intake pressures.