Resumen de Granular flow characterisation towards an enhanced bulk feeder selection
Joel Torres Serra
In the packaging industry, a wide range of powders and bulk solids are handled in bagging equipment using different bulk feeding techniques. Quality of the packaging is critically affected by feeder performance, which is conditioned by the flowability of granular materials. Therefore, selecting the adequate feeder for every application is still a challenging industrial issue, commonly approached using a combination of conventional material tests and engineering know-how. The main goal of this thesis is to develop a testing protocol suited to better assess the flowability of industrial granular materials and to inform a methodology to objectify the decision-making process of feeder selection. The new TMI granular flow tester (GFT) is presented, consisting of a quasi-two-dimensional granular column collapse set-up, which is fully-instrumented to observe granular flow phenomena. The GFT covers the aeration and deaeration of the granular columns by packing state pre-conditioning, surface morphology of the initial granular columns and the final deposit configurations at run-out, basal load profiles of material deposition for the flow stages after collapse, and the flow front propagation and near-wall kinematics by particle image velocimetry analysis. This protocol is used to define new relevant flow parameters for industrial granular materials with sizes typically ranging from the order of micrometres to a few millimetres. A data analysis methodology is proposed to verify the feeder-type classification of granular materials in industrial practice by cluster analysis on data sets of material properties describing their flowability. Using conventional properties, granular materials are clustered into as many groups as main bulk feeder classes but a significant mismatch is noted, attributed to common industrial practice and incomplete flow characterisation. For this reason, the relevant flow parameters are included in an extended property data set, notably improving detection of similarly flowing materials. Furthermore, granular flow problems are modelled by the discrete element method to include the effects of material properties at the particle level (grain shape and size, segregation, and hygroscopicity issues) in the flow behaviour observed at the bulk level. Simulation results for the GFT set-up and an industrial case study of hopper filling are incorporated into this thesis.
