Desenvolvimento e estudo da influência de um modelo de força de lubrificação em um leito fluidizado líquido utilizando CFD-DEM não resolvido

Abstract

Solid-fluid multiphase flows are common occurrences in chemical engineering applications. Due to the scale and unpredictability of the fundamental phenomena in these systems, microscopic interactions between individual particles and the surrounding fluid are not easy to study empirically. In this context, computational engineering tools like CFD-DEM (Computational Fluid Dynamics and Discrete Element Modelling) can offer a better understanding of such interactions due to its microscale approach. Most studies that employ CFD-DEM however focus on low viscosity fluids, such as gases, due to the increase in complexity that arises from interphase viscous forces with fluids like liquids. Because of this, little is known in terms of the influence of the lubrication force on the dynamics of concentrated solid-liquid flows, a dissipative force that increases the energy lost on particle-particle collisions. The present work aimed to develop, implement and validate an unresolved lubrication force model in CFD-DEM simulations, while also verifying its influence on chaotic solid-liquid flows such as a fluidized bed. Initially, a lubrication force model was implemented following an approach commonly found in literature and compared to experimental data from other authors. The model was then modified based on the results and validated successfully. Experimental fluidization data was collected for 1 mm alumina particles in water, in a fluidized bed with dimensions of 5 cm in diameter and 30 cm in length. The experimental setup was then simulated using an open-source unresolved CFD-DEM framework without the lubrication force. Once the simulation was successfully validated, they were repeated with the inclusion of the lubrication force model and results were compared between both sets of simulations but no significant change was observed. Fluidization conditions were then extrapolated by using a glycerin solution instead of water, thus increasing the fluid’s viscosity. In this case, the lubrication force caused a reduction in bed expansion and porosity, and results were closer to those predicted by empirical correlations than those produced by simulations without the lubrication force model. Overall, the results show that unresolved CFD-DEM is capable of correctly predicting the behavior of complex concentrated flows, and can be used to aid in designing liquid-solid fluidized beds.