Análise comparativa de modelos de fechamento na simulação fluidodinâmica de uma coluna de bolhas

Abstract

Bubble columns are widely used in chemical, petrochemical, biochemical, and environmental processes due to their high mass and heat transfer efficiency. In this context, gas holdup is one of the main hydrodynamic parameters, directly influencing reactor performance. This work evaluated the influence of different closure models on the prediction of gas holdup in a bubble column using Computational Fluid Dynamics (CFD) simulations performed in the STARCCM+ software. Initially, a mesh independence study was conducted using the Grid Convergence Index (GCI) to ensure the numerical reliability of the simulations. Subsequently, different turbulence models, bubble diameter models, interfacial force models, and the inclusion of bubble breakup and coalescence mechanisms were investigated. The simulation results were mainly assessed by comparison with gas holdup correlations available in the literature and were additionally confronted with experimental data. The mesh independence study confirmed numerical convergence and the reliability of the adopted discretization. Among the evaluated models, the k-ω turbulence model showed the best performance in predicting gas holdup, presenting the lowest residuals throughout the simulations. Bubble diameter correlations and drag models were found to be the most influential factors in gas holdup prediction, leading to the largest variations in the simulated results, whereas lift and bubble breakup and coalescence models had a smaller impact on this variable. The inclusion of breakup and coalescence models did not lead to consistent improvements in gas holdup prediction, despite the associated increase in computational cost. Furthermore, empirical correlations showed accuracy dependence on the operating range when compared with experimental data, failing to adequately represent all studied conditions. Overall, the results demonstrate that the appropriate selection of closure models is essential for reliable predictions of gas holdup and system hydrodynamics, confirming the potential of CFD as a tool for analysis and support in the development and optimization of bubble columns.