Estudo de modelos de perda de carga em coluna de leito fixo para sistemas gasosos

Abstract

This dissertation has the aim to test alternative models to the Ergun s model for predicting pressure drop during the flow of gaseous currents in packed beds. In this context, this work was initiated with the development of the momentum balance in its general format, simplifying the terms on the basis of other researches available in the literature. It was obtained as a result that the pressure drop in fixed column beds can be estimated by making use of friction factor equations for porous media, such as Ergun s model. In the specific case of carbon dioxide adsorption columns, a bibliographic review made for Shafeeyan et al. (2014) from last three decades revealed that 30% of research studies regarded Ergun s model and 56% neglect this phenomenon. Therefore, it was investigated 26 other alternative models to Ergun s that would correlate characteristics of the bed and the flow, such as porosity, particle shape, influence of the column walls, gaseous current velocity, density and viscosity. It was also investigated predictive models capable of providing good estimates of properties for transport of pure gases and mixtures, composed by polar and apolar molecules. It was confirmed that Peng-Robinson s model is effective in the determination of density, Lucas model is in the estimate of viscosity of pure compounds and Chung et al. s for apolar gaseous mixtures. It was performed a simulation, by means of a case study available in the literature, the pressure drop during the flow of the pure gases, nitrogen, oxygen and helium, and a gaseous mixture composed by nitrogen and oxygen with various compositions. The properties of transport were simulated by Peng-Robinson s and Chung et al. s models. The relative average deviations between experimental and simulated data for pressure drop have demonstrated the potentiality of Rose and Rizk s model in predicting experimental data, which generates better estimates than Ergun s model.