Characteristic of ligament in centrifugal granulation by spinning disc

Junjun WU,Hong WANG,Xun ZHU,Qiang LIAO,Jun LI,Lin LIN

DOI: https://doi.org/10.11949/j.issn.0438-1157.20150008

2015-01-01

Abstract:Aiming at the granulation of molten blast furnace slag by spinning disc, the visualization granulation experiment was conducted using water as the substance. The granulation process including film waving and ligament break-up was captured by high-speed camera. A self-compiled MATLAB program was employed to extract the data from the images. The ligament formation and the process of ligament break-up into droplets during centrifugal granulation were analyzed. Variation of ligament formation and droplet formation with operating conditions was investigated. The qualitative relationship between ligament formation and droplet formation was obtained. The impacts of Weber and Reynolds numbers on ligament number, ligament break-up length, droplets diameter distribution and average diameter were discussed. The results revealed that the ligaments were originated from surface instable waves of liquid film on the spinning disc and the droplets were formed by ligament break-up due to symmetric shocking waves spreading from the ends to the middle of ligament. The number of ligament increased with increasing rotary speed and the distribution of droplets became more highly concentrated as well, while the break-up length of ligaments and the average diameter of droplets were decreased as rotary speed increased. With increasing liquid flow rate, the number of ligaments increased, while the distribution of droplets became less concentrated. The break-up length of ligaments and average diameter were increased with higher liquid flow rate. Besides, larger ligament number or shorter ligament break-up length resulted in smaller droplets. Meanwhile, both of the Weber and Reynolds numbers had significant influence on the ligament formation, while only the Weber number had significant effect on droplet formation. By fitting the experimental data, correlations have been obtained to predict the granulation variation with different operational conditions quantitatively.