Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.505425
Title: Predictive model of fluidized bed melt granulation
Author: Chua, Kel Win
Awarding Body: The University of Sheffield
Current Institution: University of Sheffield
Date of Award: 2009
Availability of Full Text:
Access through EThOS:
Abstract:
This thesis describes the development of a computer model that predicts aggregation behavior in fluidized bed melt granulation (FBMG) processes. In FBMG processes, the molten binder enters at one or more points and is then distributed around the bed, these processes are necessarily spatially inhomogeneous. Current granulation modelling using Population Balance Equations (PBE) adopts a lumped approach, using a single lumped rate constant. This thesis shows how to take account of these spatially distributed processes and then predicted the apparent lumped rate constant. The first half of this thesis presents a series of experimentally validated sub-models to describe time scales of events leading to aggregation. These events are granule-granule collision, droplet-granule collision, binder spreading and binder solidification. Computational Fluid Dynamic using a twin-fluid model is used to calculate granulegranule collision time scale and droplet-granule collision time scales. A Volume of Fluid method is used to calculate binder spreading time scales. A dynamic energy balance is used to calculate binder solidification time scales. Validation of the twin-fluid model is done by comparing fluidized bed pressure time series with experiment. Validation of spreading model is done using high speed images of binder droplet spreading. Validation of solidification model is done using high speed infrared images of binder droplet solidification on glass plate. Generalizations for binder spreading time scale and binder solidification time scale are made, so that these time scales for other FBMG process can be calculated directly, reducing the reliance on computational modelling. Granule-granule collision time scale per particle is in the region of O.Ols. Binder spreading time scale is less than O.OOls~ Binder solidification time scale varies from 0.005s to O.Ols. This wider variation for solidification time scale is caused by the effect of fluidized bed operating temperature. By separation of time scales, we evaluate probability that a granule is wet upon granulegranule collision. Binder contact angle determines the wetted area covered by single droplet. Droplet solidification time scale determines the lifetime of the wetted area. This probability is used to calculate granulation efficiency. Applying granulation efficiency with collision rates allows aggregation rate constants for different process variables in FBMG to be evaluated. Our evaluation found that most aggregation takes place within 5cm of the nozzle. In general the calculated aggregation rate constant is higher than the experimentally measured aggregation rate constant. \Ve attributed the unknown effects of averaging across granule SIze, the probability of liquid bridge survival and probability of solid bridge survival with an apparent probability of survival.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.505425  DOI: Not available
Share: