The work presented in this thesis is a first step towards a new methodology to design and operate continuous granulators. Ennis and Litster (in: R.H. Perry and D.W. Green, Perry’s Chemical Engineers’ Handbook (7-th Ed.), Me Graw Hill (1997) 56-85) showed that the granulation process can be divided into three rate processes, these being wetting and nucleation, consolidation and growth and breakage and attrition. Current granulation processes traditionally consist of one granulator volume. All granulation rate processes coexist within this volume and cannot be distinguished individually. By physically separating the granulation rate processes, the design, operation and control o f each rate process can be targeted individually. The presented granulator design and operation methodology is based on the physical separation of the granulation rate processes. Mathematical tools have been developed to investigate, design and operate the wetting and nucleation, and consolidation and growth rate processes.
A nucleation model is developed to assist in the design and operation of the nucleation section of a granulation process. The nucleation model predicts nuclei size distributions as a function o f all the relevant process parameters in nucleation processes where binder liquid is added with a spray nozzle. These parameters are the binder liquid mass distribution in the spray zone, the drop size distribution of the binder liquid, the powder bed velocity through the spray zone and the nucleation kinetics represented by the nucleation ratio between binder liquid droplets and nuclei. The nucleation model simulates a real spray of binder liquid droplets landing on a powder bed surface passing once through the spray zone. The model then determines which droplets overlap at the powder bed surface, taking into account the size difference between the original droplets and the formed nuclei. When two or more nuclei overlap, the model forms a single larger nucleus from the individual nuclei. By repeating this procedure for all binder liquid droplets sprayed, a nuclei size distribution is obtained. Variation of the model parameters showed that the increasing the binder liquid flow rate or decreasing the powder bed velocity have the same effects on the simulated nuclei size distributions. Both variations widen the nuclei size distribution towards larger nuclei sizes due to an increased extend of droplet overlap. An increase of binder liquid drop size at constant binder liquid flow rate shifts the distribution towards smaller nuclei due to a lesser extent of droplet overlap. Small increases of the nucleation ratio between binder liquid droplets and formed nuclei showed the largest influences on the nuclei size distributions of all operating parameters tested, due to changes of the total volume of nuclei being produced.