Markov chain modeling of fluidized bed granulation incorporating simultaneous aggregation and breakage

Fluidized bed granulation to achieve particle size enlargement is a significant process in chemical engineering. A key aim of the modeling of such a system is prediction of the evolution of the particle size distribution with time. The traditional tool to achieve this is through the use of population balances. This Article proposes an alternative modeling approach of Markov chain, which is a well-known tool in stochastic theory. Specifically, in this Article, it is employed to model bottom spray Wurster type fluidized bed granulation and a one-dimensional chain where the particle size is the state variable was used. Three important elements of Markov chains are the initial state vector a(t), the transition time step τ, and the transition matrix P. The transition matrix was assembled from consideration of breakage and aggregation dynamics. Experiments were conducted to validate the approach using glass beads as the elementary particles and polyethylene glycol as the liquid binder. From the results, it is shown that the Markov chain method is an efficient and accurate tool to analyze the particle size evolution of a size enlargement process.