Agglomeration discrete modeling

One manner of using the presented results is to incorporate them in the traditional way of tackling fluid bed granulation theoretically, namely population balance modeling. This can be achieved by expanding the population balance to more internal coordinates than just particle size (see Volume 1 of this series. Chapter 6, Section 6.9.1). The additional property in the case of the present example would be wet a lomerate composition, defined either by the mass fraction of solids within one particle or the binder/solid ratio. The latter can be further spht up to account for the spatial distribution - and, thus, accessibUity - of the hquid binder and for the thickness of the binder layer on the outer surface of the agglomerate. Alternatively, discrete models of agglomeration (see Section 7.7) could be expanded to account for non-spherical primary particles. 

Kapur and Fuerstenau (K6) have presented a discrete size model for the growth of the agglomerates by the random coalescence mechanism, which invariably predominates in the nuclei and transition growth regions. The basic postulates of their model are that the granules are well mixed and the collision frequency and the probability of coalescence are independent of size. The concentration of the pellets is more or less fixed by the packing... 

Fig. 16.3 (a) Schematic representation of the spherical catalyst agglomerate model, (b) Modified agglomerate model with distribution of discrete Pt particles. Reproduced from [82] with permission of The Electrochemical Society... 

Cetinbas EC, Advani SG, Prasad AK (2013) A modified agglomerate model with discrete catalyst particles for the PEM fuel cell catalyst layer. J Electrochem Soc 160 F750-F756... 

Drying and liquid penetration are also important for the process already discussed in Section 7.3.3, namely spray fluidized bed agglomeration. The reason for this is that agglomeration takes place with the help of droplets sprayed on the particles, so that it slows down when such droplets are lost either by evaporation (drying) or by liquid penetration into the porous substrate. Influences of this kind can be captured very well with the help of respective micro-scale models integrated into discrete simulations, as we will see in Section 7.7. 

The coal particles can be tracked as parcels in an Eulerian-Lagrangian framework. Discrete phase model (DPM) are used to define the injected particles that enter the reactor. In the case of INCI, simulation values for axial velocity of -1.732 m/s and a radial velocity of -l.Om/s of the particles must be provided. If agglomeration is neglected, a maximum particle diameter of 0.1 mm, a mean diameter of 0.09 mm, and a minimum diameter of 0.001 mm are assumed according to a Rosin-Rammler-Sperling-Bennett distribution with a spread parameter of = 0.688 in 10 individual groups for fluid-bed coal (see also Section 3.12.3.3). Particles can be treated as nonspherical with a shape factor of 0.85. 

---