Measurements of granular flow dynamics with high speed digital images

Publisher Website
Google Scholar
Abstract
The flow of granular materials is common to many industrial processes. This dissertation suggests and validates image processing algorithms applied to high speed digital images to measure the dynamics (velocity, temperature and volume fraction) of dry granular solids flowing down an inclined chute under the action of gravity. Glass and acrylic particles have been used as granular solids in the experiment. One technique utilizes block matching for spatially averaged velocity measurements of the glass particles. This technique is compared with the velocity measurement using an optic probe which is a conventional granular flow velocity measurement device. The other technique for measuring the velocities of individual acrylic particles is developed with correspondence using a Hopfield network. This technique first locates the positions of particles with pattern recognition techniques, followed by a clustering technique, which produces point patterns. Also, several techniques are compared for particle recognition: synthetic discriminant function (SDF), minimum average correlation energy (MACE) filter, modified minimum average correlation energy (MMACE) filter and variance normalized correlation. The author proposes an MMACE filter which improves generalization of the MACE filter by adjusting the amount of averaged spectrum of training images in the spectrum whitening stages of the MACE filter. Variance normalized correlationmore » is applied to measure the velocity and temperature of flowing glass particles down the inclined chute. The measurements are taken for the steady and wavy flow and qualitatively compared with a theoretical model of granular flow.« less

This publication has 0 references indexed in Scilit: