Mathematical Model for Agglomeration Process of Milk Powder

Abstract

Stickiness during milk spray drying can lead to the agglomeration of milk powder and damage the processing equipment. A mathematical model can achieve a better understanding. In this work, the Distinct Element Method (DEM) simultaneously with Computational Fluid Dynamics (CFD) was used to describe skim milk powder's agglomeration process. The study comprised 2 parts: surface stickiness mechanism and agglomeration of sticky powder. Start with particle formation, the droplet size, and the number of particles produced can be calculated and used to predict the droplet's surface stickiness. These reveal the effect of moisture content, droplet surface temperature, droplet size after drying, and sticky point temperature. Then, the agglomeration of sticky powder inside the spray chamber was predicted. Besides, the particle and fluid motion inside the spray chamber were also determined. Then, the particle size distribution after agglomeration was obtained. Furthermore, parts of the model were validated with the experimental data of Williams et al. (2009), which has three different droplet sizes, 56.8, 78.28, and 108.5 micrometers. The results gave the same trend as the sticky surface of the powder. The droplet's moisture contents rapidly decreased in the first period and fell to a critical value, which was 0.044, 0.048, and 0.061 kg water/kg solid, respectively. The periods of a sticky surface were around 0.033, 0.03, and 0.024 seconds. The largest droplet size was selected for the study of the agglomeration process. This model could predict the agglomeration of sticky powder since there were 216 from 900 droplets agglomerated. Moreover, the largest droplet size was 100.6 micrometers, and the most popular was 79.9 micrometers, which were the size of the un-agglomerated powder.