Fast Methods to Analyze High-Speed Images of HCCI and Spark-Assisted HCCI Ignition Events

Abstract

This study presents two new methods to rapidly analyze the ignition and combustion events recorded during in-cylinder high-speed imaging of homogeneous charge compression ignition (HCCI) and spark-assisted (SA) HCCI engine operation. Two categories of ignition have been observed during HCCI and SA HCCI: auto-ignition of parcels of fuel/air and ignition and propagation of reaction fronts. In this work, two methods of image analysis are developed which accommodate differences in the ignition and propagation events for the two categories. The methods are described by application to high speed imaging data acquired from an optically accessible engine operated in HCCI and SA HCCI modes using indolene fuel. Using method I, the projected area of reaction is identified in each frame, and a circle with a characteristic radius that yields the equivalent area is defined. The rate of area expansion is used to define the speed of reaction. Using method II, an edge-finding algorithm is used to identify the location of well-defined reaction fronts. The maximum dimensions of the reaction front in an x-y coordinate frame are measured. The projected area based on the average of these maximum dimensions is then determined. The change in the characteristic radius of the projected area is defined as the average propagation speed for the reaction front. The two analytical methods are applied to typical HCCI and SA HCCI image sequences. Outcomes of the image analysis indicate propagation speeds are typically between 2–5 m/s prior to volumetric ignition. Volumetric ignition is identified by propagation speeds greater than 20 m/s. The imaging data also identify times when reaction quenching occurred after spark ignition as times when negative propagation speeds are observed. When applied to the same imaging sequence, the two methods yield consistent results for propagation speeds; however, the method based on edge-finding exhibits larger variance.