Engine tests were conducted to study the effect of fuel-air mixture preparation on the combustion and emission performance of a two-stroke direct-injection engine. The tests were performed with a wide range of injection timings for several injection systems in order to create a variety of fuel-air mixtures. The experiments were performed on a single cylinder two-stroke water-cooled loop-scavenged research engine using an intake manifold injection system to deliver a nearly homogeneous air-fuel mixture, an air-assist injector with a highly atomized spray, and a single-fluid high pressure swirl injector with a dense penetrating spray. The combustion chamber geometry was not optimized for any of the injection systems, but it was confirmed that each injection system could operate over the entire speed/load range of the engine.

The engine experiments were performed at an overall A/F ratio of 30:1, where stratification was necessary to ensure stable combustion; and A/F ratios of 20:1 and 15:1 where it was possible to operate in a near homogeneous mode. Injection timing windows of stable combustion were found for both the air-assist and high pressure swirl injectors. These windows occurred for very retarded injection timings (highly stratified mixtures) for certain operating conditions, and very advanced injection timings (nearly homogeneous mixtures) for certain operating conditions.

Several emission formation mechanisms for direct-injection two-stroke stratified-charge combustion were presented and examples of these mechanisms were identified based on the engine test results. For all cases tested, the HC emissions from the high pressure swirl injector were dramatically higher than from the air-assist injector. It was believed that the higher HC emissions were caused by large drops from the leading edge of the spray and at the end of injection that would be difficult to vaporize with the cylinder, the high penetration rate of the high pressure swirl injector which may have increased short-circuiting of unburned fuel or may have wetted the piston, or the fact that the air-assist injector produced smaller droplets and contained oxygen within the spray which would have created a leaner burning mixture resulting in lower HC emissions.

For stratified operating conditions, it was found that the injection timing affected the local burning zone air-fuel ratio, and thus strongly influenced the HC and CO emissions. Injection timings retarded from the optimum produced both increased HC and CO emissions because the fuel spray had less time to mix with the cylinder gases, resulting in a richer burning mixture. For injection timing advanced from the optimum, there was increased HC emissions without a corresponding increase in CO emissions which is consistent with overmixing of the fuel spray. Experiments performed using nitrogen instead of air in the air-assist injector confirmed the importance of the burning zone A/F ratio on the emission performance of direct-injection stratified-charge combustion.