Spray drying experiments and CFD simulation of guava juice formulation

Abstract

This paper presents a computational fluid dynamics (CFD) approach to model the spray-drying of a formulation of guava juice. The computational model incorporates the characteristic drying rate curve (CDC) of the formulation. The CDC is derived from the drying history of single droplets suspended on a thin glass filament. The computational domain is based on the geometry of a small production spray-dryer with a capacity of 10 kg/h. A two-phase Eulerian-Lagrangian model implemented in a commercial CFD package (ANSYS-FLUENT) is used to simulate the gas-particle interaction in the dryer. Hot-wire anemometry (HWA) measurements of the air flow in the drying chamber are used to calibrate the CFD model. The URANS k-ω SST and the Scale-Adapted Simulation turbulence models are employed to compute the unsteady three-dimensional flow field. Model calibration suggests that the swirl number of the air flow entering the dryer is about 0.61. Simulations predict high axial velocities along the center-line and a narrow jet core in the upper part of the dryer. Lagrangian particle tracking is used to calculate the particle trajectories starting from the perimeter of the rotary disk atomizer and ending at the dryer walls and bottom outlet. Simulation results of particle size distribution, time-average temperature and humidity profiles are presented in order to analyze the overall drying performance. The procedure shown in this work emphasizes the need of implementing drying kinetics with experimental support, and suitable inlet flow conditions to perform detailed CFD simulations of the spray-drying of fruit juices.