Injection Timing and Bowl Configuration Effects on In-Cylinder Particle Mass 921646

The formation of particles in the combustion chamber of a direct injection diesel engine has been studied with the use of the Total Cylinder Sampling Method. With this method, nearly the entire contents of the cylinder of an operating diesel engine can be quickly removed at various times during the combustion process. The particle mass and size distributions present in the sample can then be analyzed. If quenching of the combustion process is quick and complete, the resulting samples are representative of the particle mass and size distributions present in the cylinder near the time sampling begins. This paper discusses the effect of injection timing and piston bowl shape on the particle formation and oxidation. Example size distribution measurements are also shown.

The particle concentrations in the cylinder were measured for three different injection timings with the standard piston installed in the engine. Measurements were also made at the base timing with an alternate piston containing a more reentrant bowl. The particle concentrations rise quickly after the start of combustion, peaking approximately 12 degrees later. Peak levels are 4 to 9 times the exhaust levels. Oxidation during the expansion stroke reduces the particle mass concentration to the exhaust levels. Injection timing had a significant effect on the profiles measured. While peak concentrations were similar, the oxidation was less effective in reducing the mass as the timing was retarded. This resulted in the advanced timing condition having the lowest exhaust concentrations of the standard piston conditions. At other conditions, oxidation was less effective in reducing the mass, resulting in higher exhaust concentrations.

The reentrant bowl configuration results showed peak cylinder concentrations similar to the standard piston, but the profile was broader than for the standard bowl. Oxidation continued further into the expansion stroke, leading to exhaust concentrations only half as high as for the standard bowl configuration. Measurements of both the agglomerate and primary particle size distributions can also be made on samples collected during these experiments. This paper presents an example of these measurements. The size of the agglomerates appears to approach that of the exhaust by approximately 20 CAD after the start of combustion. The size of the primary particles increases during early combustion but is reduced later in the cycle due to oxidation. The total reduction in the primary particle size is not enough to account for the reduction of total particle mass, indicating that some particles are being completely oxidized while others are oxidized only partially.