A Comparative Analysis of Direct Injection into a Pressurized Chamber Using an Automatic Image Treatment Methodology
A multi-hole direct injection injector was studied by means of image analysis. Methodologies based on an automatic process of cone angle measurement and edge detection were applied for the spray images generated by a 100 bar injection pressure discharged into a pressurized rigid chamber. A criterion based on pixel values was taken to localize the spray edges as angular coordinates and also with x and y position data. The high pixel values were associated with liquid phase while the low pixel values were associated to its absence. Computational codes written in MATLAB environment were used to analyze the numerical matrices associated to the images. Using the written MATLAB codes, a comparison of the effect of atmospheric back pressure, inside the chamber, on the spray pattern, cone angle and spray penetration were evaluated. The chamber was pressurized with 2.5, 5.0, 7.5 and 10 bar of back pressure. The tested fluid injected was EXXSOL D60 for simulating ethanol fuel behavior.