Assessment of Premoli et al. Correlation Predictions in Horizontal Stratified-Wavy Two-Phase Flow Encountered at Small Diameter T-Junction

Georgios K. Makrygiannis, Dionissios P. Margaris

CFD model, two-phase flow, void fraction, Premoli et al. correlation

Two-phase flow is present in many and of different kind of industrial applications, especially in the form of gas-liquid two-phase flow. Despite of its common occurrence, its understanding is quite limited in comparison to single-phase flow due to its complexity. One critical parameter to characterize a gas-liquid two-phase flow and thereby to appropriate size the required industrial equipment is void fraction. The latter is affected by flow rates, fluid properties and flow gradient, whereas in turn it influences other important flow parameters such as pressure drop and heat transfer among others. Though there is available a plethora of void fraction correlations, most of them are limited to be accurate for similar fluid flow problems to that developed for, and thus they cannot be considered suitable to be utilised irrespective of the applied inlet conditions. However, a slip ratio correlation proposed by Premoli et al. has been proven quite accurate in void fraction predictions, and thus it is recommended to be used in case of horizontal and upward inclined pipes regardless the established flow regime, whereas its implementation does not require complex or iterative procedure to be performed. This hypothesis was evaluated by conducting numerical analysis and comparing the obtained results for void fraction, average velocity, and two-phase mixture density with the corresponding predictions of Premoli et al. correlation. It was found that in general Premoli et al. correlation over predicts the void fraction and average water velocity values, while under predicts the average air velocity and two-phase mixture density. However, the observed error does not indicate that the adoption of Premoli et al. correlation is a mistaken option a priori with the final choice of its usage being determined by the desired accuracy required from the application.

https://doi.org/10.31871/IJNTR.6.3.8