Experimental and CFD evaluation of bubble diameter and turbulence model influence on nonlinear flow dynamics in vertical columns: A comparative study

Abstract

This study examined the impact of a few numerical factors, such as mesh density, drag model selections, turbulence models, and bubble size descriptions on the numerical accuracy of gas-liquid (GL) flow. Multiphase flow in a vertical pipe with a diameter of 76.2 mm was simulated using the Eulerian-Eulerian (EE) model provided by ANSYS Fluent. A comparison with experimental findings shows that the two-phase flows are significantly impacted by the bubble size. To determine their effect on simulation accuracy, various kinds of turbulence models are assessed such as Standard $k-$epsilon, RNG $k-$epsilon and the Realizable $k-$epsilon models. Four superficial gas velocities were used to investigate turbulence models where superficial liquid velocity is constant. Realizable $k-$epsilon model often offers the closest fit to the experimental data. Particularly at high gas velocities and improved bubble sizes. This model often gives a more perfect interpretation of the void fraction and phase distribution than the other two models. Its improved performance implies that the Realizable $k-$epsilon model is well managing anisotropy in the turbulence, enabling more precise predictions of the void percentage in intricate, extremely turbulent flow regimes. To enhance the precision of nonlinear flow predicts, future research should concentrate on investigating complex turbulence models and higher-order numerical schemes. Additionally, examining the effects of bubble coalescence, breakup dynamics, and three-dimensional flow effects would allow for a deeper comprehension of flow behaviour in vertical columns.