No abstract

Experimental study has been conducted on auto ignition and knocking phenomena of stratified mixture using a rapid compression machine (RCM) in order to investigate the effects of fuel concentration gradient on the auto ignition and combustion characteristics in the condition that mean equivalence ratio is lean. N-heptane is used as a fuel and mean equivalence ratio is 0.6, 0.8 and 1.0 in the combustion chamber. In the chamber, the lower the vertical location is, the richer the concentration of the mixture is. The difference of the equivalence ratio from bottom to top in the chamber is varied from 0 to 1.4. Mixture has no gradient in horizontal direction. Initial temperature of the mixture is 290K and pressure is 0.1MPa before compression. The diameter of the combustion chamber is 65mm and the compression ratio is 10.5. High-speed video camera is set in front of chamber so that direct images of mixture from cylinder axis direction can be obtained. A pressure transducer is used to obtain pressure histories of chamber, from which ignition delay, pressure rising rate and knocking intensity are determined. The results show that: 1. Flame speed of stratified mixture obtained from direct flame images is much faster than that of calculated laminar burning velocity for corresponding mixture, thus, rapid spread of flame in the experiment is caused not by flame propagation but by consecutive auto ignition. 2. Ignition delays of stratified mixture, not depending on the gradient of mixture, are constant as far as mean equivalence ratio is same and decrease with the decrease of mean equivalence ratio. 3. Pressure rising rate of stratified mixture combustion increases with the decrease of the gradient of mixture at 0.4≤φ≤1.0. 4. In homogeneous condition, knocking intensity of smaller mean equivalence ratio is smaller. 5. With slight gradient, knocking intensities of same mean equivalence ratio are similar to that of homogeneous mixture. In excess of certain value of the gradient, knocking intensity is smaller as the gradient is larger.

No abstract