Estudo experimetal e numérico da dinâmica de grãos de café em um tambor rotatório

Abstract

The study of particle dynamics inside a rotating drum is extremely important due to the wide range of application for this equipment in different industrial sectors. In order to produce coffee, in the drying and roasting stages, rotary drums are used. Considering the fact that Brazil is the largest producer and exporter of coffee in the world, improving the efficiency of these equipments has great relevance. Thus, the present work developed an experimental and numerical study about the dynamics of coffee beans, under green and roasted conditions, inside a rotating drum. Experimentally, for both green and roasted coffees, the transitions between rolling, cascating, cataracting and centrifuging regimes were evaluated using an image analysis technique. The different density values in both conditions, green and roasted, significantly influenced the cataract-centrifuging transition. Further analyses were made regarding the rolling regime: velocity profiles of the particles were evaluated as well as the thickness of the active layer for different filling degrees and rotation speeds of the drum. This thickness was observed as dependent on both variables. It was found great similarity between the behavior of the active region for coffee beans before and after the roasting process, considering a constant filling level. About the numerical study, two methodologies were employed: Lagrangian and Eulerian. For the Eulerian approach, two boundary conditions on the wall were tested (specularity coefficient and no-slip ) and the model with non-slip condition coupled with frictional viscosity parameter proved to be an efficient model predicting the velocity profiles for green coffee beans in the rolling regime. As for the Lagrangean approach, a calibration was performed involving the coefficients of rolling and static friction (particle-particle and particle wall). These parameters were evaluated using a Central Composite Experimental Design and it was also possible to find the best combination between them, capable of reproducing the behavior of the green grains for particle velocity profiles.