Control of fluidized bed coating particles using Gaussian spectral pressure distribution

Fluidization has been used extensively in many industrial processes in chemical, food and pharmaceutical branches. However, the operational conditions of these processes frequently lead to the defluidization phenomenon or to the total collapse of the particles in the bed. The maintenance of stable conditions in the fluidization regime during fluidized bed coating processes is very important, because the moisture content excess can cause the defluidization of the bed. The objective of this work was to apply a new methodology, known as Gaussian spectral pressure distribution, to monitor and control the defluidization phenomenon in a fluidized bed coating process using microcrystalline cellulose as fluidizing particles. The work was performed in two stages: 1) monitoring of the fluidization regimes during the development of the coating process without control and 2) control of the airflow rate and of the coating suspension flow rate using PI controllers. The experiments were carried out varying solid particle mass, coating suspension flow rate and excess air velocity in relation to the minimum fluidization velocity working with a temperature of 70 degrees C. The Gaussian mean frequency evolution showed the fluidization regime transitions and it allowed to define a band of stable regime (6.0 Hz to 7.0 Hz), which was used as a set-point range of the controllers to manipulate the signal of command for frequency converter and for the peristaltic pump. The application of Astrom and Hagglund [29] method allowed obtaining the PI controller parameters for the converter and the reaction curve method provided an initial guess of the PI controller parameters for the pump. The experiments of microcrystalline cellulose coating in closed-loop showed that the use of a control system allowed obtaining better fluid-dynamic conditions of the bed in relation to the process without control. (C) 2011 Elsevier B.V. All rights reserved.