Effects of auxiliary injection on diesel engine combustion

A two-stroke diesel engine was outfitted for operation with an electronic solenoid-controlled unit injector and an additional solenoid-controlled air-assisted injector at the inlet ports. Factorial experiments were designed in order to quantify effects of pilot (or 'split') and port auxiliary injection on main fuel combustion. Access through an existing pressure transducer port allowed installation of a sapphire window to the combustion chamber with very little disturbance to the combustion system. A coherent fiber-optic bundle permitted remote visualization of the combustion event. Use of a gateable intensified solid-state camera permitted imaging at high effective shutter speeds at arbitrary times in the engine cycle. Imaging and two-color temperature and soot concentrations measurements were performed.; Results indicated that interactions between the experimental parameters as well as their main effects are important in analyzing auxiliary injection. For an example of main effects only, the bulk gas temperature at main injection was primarily influenced by pilot or port fuel quantities. Conversely, analysis of indicated specific fuel consumption and peak pressure involved interaction of main and auxiliary injection terms in both cases.; Other results indicated that addition of a pilot injection could effectively separate the premixed and diffusion portions of the combustion process. The port fuel was observed to undergo a two-stage ignition process and suppress the premixed burn of the main fuel. The early premixed burn spike of the pilot fuel could increase the bulk gas temperature at main injection by over 350 K. The two-stage ignition process of the port fuel could cause an increase in this quantity of up to 200 K.; Imaging results indicated a low-intensity diffuse ignition, away from the injector tip, for both the pilot spray in pilot-main tests and the main spray in the main-only runs. Remnants of the burning pilot spray congregated near the injector tip where a region of flame remained until main injection arrived. The main fuel ignited upon injection into this region.; Temperature measurements indicated that pilot injection resulted in a region of 2300 K or greater flame temperature around the injector nozzle at the start of main injection. The two-stage ignition process of the port fuel was not detectable with this measurement system. Small-scale variations on the order of 100 K were evident in most temperature images. The soot concentration in the pilot flame did not increase appreciably during the pilot-main delay period.