Liquid atomisation processes in the crankcase of an IC engine: a computational and experimental study

Abstract

The present thesis explores the mechanisms of aerodynamic breakup, jet disintegration and droplet formation inside the crankcase of an internal combustion engine. The work employs advanced experimental and computational tools, which allowed the description and detailed understanding of the flow processes driving liquid atomisation. The literature review identified numerous types of droplet-flow interactions, flow structures and mechanisms which required further investigation. These important parameters are currently ignored in most engineering design. Droplets, upon generation in the crankcase, is shown to be far from spherical, an assumption usually made in the engineering practice. The effect of droplet non-sphericity on its shape evolution and disintegration process was a primary research objective. The experimental part of the work allowed the detailed investigation of droplet-crossflow interactions and its subsequent atomisation for a range of Weber, Ohnesorge numbers and initial droplet aspect ratios. New, equivalent Weber and Ohnesorge numbers, incorporating the initial droplet non-sphericity, are proposed, which are able to classify the breakup of spherical and non-spherical droplets under the morphological classification charts in the literature. Additionally, new empirical correlations are presented for the breakup initiation time and maximum cross-stream droplet diameter. The experiments were further clarified with computational investigations as regards the breakup initiation time, droplet shape evolution and transitional droplet drag coefficient. The computational investigation studied the strength and location of the rollex, which for topological reasons must exist on the droplet – gas rear interface. A novel experimental setup of a rotating disc was manufactured to partially simulate the crankshaft. The jet disintegration modes and droplet formation processes, due to its rotation, are explored and described. Detailed investigation of the effect of the crankshaft rotational speed and liquid flow rate led to the clarification of the disintegration mechanisms and to the study of the breakup length of the liquid jet. Finally, a initiative was undertaken to link light-field imaging with liquid atomisation processes. A plenoptic camera prototype, designed and manufactured in-house, was used in a simplified droplet arrangement. Further to that, an algorithm was developed for the three dimensional reconstruction of the plenoptic images. The results indicated the high potential of the plenoptic concept to multiphase flows.