Discrete element method simulation of a twin screw wet granulator

Granulation is a method of particle size enlargement via agglomeration of primary particles to form larger particles. In a Twin Screw Wet Granulator (TSWG), a binder liquid is sprayed onto the surfaces of active pharmaceutical ingredient particles to promote the formation of large granules with specific morphology, flowability, compressibility and other properties. The morphology of granules formed from a TSWG process depends on the material properties of the primary particles and operating conditions applied but mechanistic understanding of the granulation process remains incomplete even today. A large number of experiments are typically performed to determine the operating parameters required for a specific type of primary particles. The Discrete Element Method (DEM) has been applied extensively by researchers for simulations of various types of particulate processes. However, good agreements between DEM simulations of the TSWG and experimental measurements have not been demonstrated in the literature to date. In this study, the DEM was coupled with a capillary liquid bridge model to simulate the granulation process in a TSWG using the EDEM software, and Python programs were applied for data collection and post-processing. Different liquid contents (30%, 45%, 60%), screw speeds (450 rpm, 900 rpm) and solid throughputs (12.5 kg/h, 25 kg/h) were applied to investigate the effects of these parameters on the residence time distribution (RTD) of particles in the TSWG and particle size distribution (PSD) of the granules formed in various regions of the granulator. It was observed that agglomeration of particles occurred in the conveying elements of the TSWG, and large granules were kneaded in the kneading elements to generate new granules with different shapes and sizes. The aspect ratio (AR) was calculated to evaluate the shapes of the granules formed. Comparisons of simulation results with experimental data reported in a previous study were conducted to validate the accuracy of the simulation results.