Análise da secagem de partículas de alumina em secador vibrofluidizado
Reis, Térbio Geraldo dos
MetadataShow full item record
Activated alumina is widely used as a desiccant or as an adsorbent. In order to reuse the material, it is necessary to submit it to the drying process. In this context, a research was carried out to obtain the alumina with the lowest possible moisture using vibrofluidized bed dryer. For a better comprehension of the drying of the material, the development of this study was divided in parts and the first one consists of kiln drying in which the only variable is the air temperature. Then, intending to analyze the variables that are capable to influence on drying process, a thin layer drying was carried out using a fixed bed dryer. In this case there are two variables: the temperature and velocity of the drying air. After that, the material was classified according to Geldart s classification (1973), which allows the inference of important data about the material fluid dynamic behavior in fluidized beds. After having the material classified, a study about the fluid dynamic of the material was performed, followed by its drying in vibrofluidized bed dryer. Tests were carried out to determine the equipment s operating range. Afterwards, the drying behavior was analyzed in different situations, in which air temperature and velocity and also vibration amplitude and frequency were changed. In data processing, diffusive, empirical and ANNs models were fitted to experimental data to analyze which one could represent alumina s drying kinetics. In the range of experimental conditions investigated, it was observed that: alumina s drying kinetics and its fluid dynamic present different behavior for the same dimensionless vibration number obtained from two different combinations of amplitude and frequency; diffusive models fitted well to experimental when the right boundary conditions were used; the empirical models which best fitted to experimental data were Page (1974) and Overhults et al. (1973) models; and a single ANN was able to fit to all the experimental data.