Fertilizers, detergents, animal feed, or food ingredients – many products of our daily life are solids produced in fluidized beds via spray granulation. During granulation fine powders are converted to granular materials by injection of a liquid, which allows the adjustment of defined properties as e.g. size or flowability. The main quality criteria of the products are a homogeneous composition and well-defined particle size, size distribution, and shape. To obtain a high productivity fluidized bed spray granulation is often operated continuously. However, the fine tuning and control of particle size at the outlet of the granulator remains a challenge, with a significant fraction of the solid that is being treated not meeting the desired criteria and requiring further costly treatment or being directly discarded.
- Development of a real-time monitoring system for the control of particle size distribution
- Design, construction and start-up of a two-chambered fluidized bed spray dryer (Fig. 1)
Particle size distribution can be controlled by real-time computer vision and the actuation of key operational parameters such as inlet flow rate or aeration modes (Fig. 2). Thus, the main objective of this project is the development of a novel real-time control framework, that us able to assess the correct actuation following the data obtained from the online particle size distribution system (PSD) and the use of novel real-time (and high resolution) rCFD simulations.
The use of a homogeneous aeration close to the minimum fluidization velocity promotes the segregation of particles with big particles sinking to the bottom of the reactor and light particles staying in the upper section of the bed. This phenomenon allows the top sprayer to affect preferentially these latter particles, increasing their size. Once particles have been coated the bed can be heterogeneously aerated in order to homogenize the solid mixture. The combination of both aeration modes allows the fine control of particle size, whose effect will be predicted through computer simulations (Fig. 3), in order to asses when each of the aeration modes must be use.
This project is fully funded by the DFG (HE 4526/31-1)
Department of Particle Flow Modelling, Johannes Keppler University, Linz, Austria