Modelagem com difusividade efetiva via analogia da lei de fick no processo de secagem de soja

Abstract

Mato Grosso is a Brazilian state that stands out for being a great soybean producer, which makes the use of techniques to reduce losses in production inherent to the poor packaging of these grains very relevant. In this respect, drying is a unitary operation that makes it possible to remove the moisture content, prolonging the life of the material, preventing its deterioration, as well as reducing its weight and volume and thus providing better conditions for storage and transportation. Thus, knowledge about the aspects of this process can guarantee greater control and precision of the specifics of the products you want to obtain. As an example, the diffusive models make it possible to predict the resistance to water removal as the variables of this process change. In this way, the present research is guided by the objective to evaluate a drying kinetics in a fixed bed of soybeans of the cultivar Glycine max (L.) Merrill, in order to evaluate a variation of the total diffusive parameter throughout the drying process. By bibliographic revision, kinetic data were collected corresponding to the drying of this seed in a thin layer. From the selected reference, models were developed and effective constant and variable diffusivities were obtained, the latter being calculated through the method of sequential adjustments of the analytical solution to the Fick's Law analogy of diffusion, considering negligible external resistance and spherical geometry. The results lead to the interpretation that there was a better adjustment from the statistical point of view of the model, obtaining the minimum adjustment conditions for R2 = 0,998 and χ² = 6.44 ∙ 10-5 , as well as the predictions of moisture content, thus having an analysis of the mass transfer phenomena during drying for the variable parameter. Thus, it was demonstrated that the methodology of sequential adjustments was equally relevant for the development of the spherical geometry model used in this work.