A Fast and Improved Tunable Aggregation Model for Stochastic Simulation of Spray Fluidized Bed Agglomeration

Abstract

Agglomeration in spray fluidized bed (SFB) is a particle growth process that improves powder properties in the chemical, pharmaceutical, and food industries. In order to analyze the underlying mechanisms behind the generation of SFB agglomerates, modeling of the growth process is essential. Morphology plays an imperative role in understanding product behavior. In the present work, the sequential tunable algorithm developed in previous studies to generate monodisperse SFB agglomerates is improved and extended to polydisperse primary particles. The improved algorithm can completely retain the given input fractal properties (fractal dimension and prefactor) for polydisperse agglomerates (with normally distributed radii of primary particles having a standard deviation of 10% from the mean value). Other morphological properties strongly agreed with the experimental SFB agglomerates. Furthermore, this tunable aggregation model is integrated into the Monte Carlo (MC) simulation. The kinetics of the overall agglomeration at various operating conditions, like binder concentration and inlet fluidized gas temperature, are investigated. The present model accurately predicts the morphological descriptors of SFB agglomerates and the overall kinetics under various operating parameters.

Modeling the agglomeration of spray-dried milk powder in a spray fluidized bed

“…The type of solid substrate is critical in determining the kinetics of the SFB agglomeration process. In general, droplet aging in SFB agglomeration of non-porous hard primary particles is solely due to convective drying of the deposited droplets [1][2][3][4][5][6]. In the case of porous hard primary particles, the aging process of the deposited liquid layer is additionally caused due to penetration (imbibition) into the pores of the substrate [7,8].…”

“…It is also assumed that the water is uniformly distributed in the puddle space, with the area of the puddle equal to the droplet cap area and the height remaining unchanged. For hard porous and non-porous solid substrates, the kinetics is modeled by the decrease in droplet height [1][2][3][4][5]. For amorphous solids, the kinetics is controlled by the amount of water evaporated from the puddle [9].…”

“…Agglomerates are reconstructed with the help of aggregation model proposed by Singh and Tsotsas [2]. It is an off-lattice particle-cluster modified polydisperse tunable aggregation (MPTSA) model.…”

“…WMP agglomerate and SMP agglomerate with 100 primary particles (𝐷 % = 200 𝜇𝑚) are generated at their respective fractal properties by means of the MPTSA model [2]. It was concluded in [12] that the porosity of the WMP agglomerates is lower by 0.05-0.15 than the porosity of the SMP agglomerates.…”

“…Morphology is essential to characterize the product formed. In order to extract the morphology of the formed agglomerates, an aggregation model from [2] is employed. The transient growth of soft porous particles is depicted using the enhanced MC model.…”