Three-dimensional particle image tracking for dilute particle-liquid flows in a pipe

X-H Pan,R Luo,X-Y Yang,H-J Yang
DOI: https://doi.org/10.1088/0957-0233/13/8/307
IF: 2.398
2002-01-01
Measurement Science and Technology
Abstract:A three-dimensional (313) particle image tracking technique was used to study the coarse spherical particle-liquid flows in a pipe. The flow images from both the front view and the normal side view, which was reflected into the front view by a mirror, were recorded with a CCD camera and digitized by a PC with an image grabber card. An image processing program was developed to enhance and segment the flow image, and then to identify the particles. Over 90% of all the particles can be identified and located from the partially overlapped particle images using the circular Hough transform. Then the 3D position of each detected particle was determined by matching its front view image to its side view image. The particle velocity was then obtained by pairing its images in successive video fields. The measurements for the spherical expanded polystyrene particle-oil flows show that the particles, like the spherical bubbles in laminar bubbly flows, tend to conglomerate near the pipe wall and to line up to form the particle clusters. As liquid velocity decreases, the particle clusters disperse and more particles are distributed in the pipe centre region.
What problem does this paper attempt to address?