The pseudo-homogeneous flow regime in large-scale bubble columns: experimental benchmark and computational fluid dynamics modeling

A precise prediction of the fluid dynamics in bubble columns is of fundamental importance to correctly design “industrial-scale” reactors. It is known that the fluid dynamics in bubble columns is related to the prevailing bubble size distribution existing in the systems. In this respect, multiphase computational fluid dynamic simulations, in the Eulerian multi-fluid framework, are able to predict the local bubble size distributions and, thus, the global fluid dynamics from the fluid flow conditions and by applying modeling closured. In particular, in in “industrial-scale” reactors, owing to the large gas sparger openings, the “pseudo-homogeneous” flow regime—characterized by a wide spectrum of bubble sizes—is typically observed. Unfortunately, reliable predictions of the “pseudo-homogeneous” flow regime are limited up to now: one important drawback concerns the selection of appropriate models for the coalescence and break-up. A set of closure relations was collected at the Helmholtz-Zentrum Dresden-Rossendorf that represents the best available knowledge. Recently, the authors have extended the validation of this set of closure relations to the “pseudo-homogeneous” flow regime, by comparing the numerical predictions to a comprehensive experimental dataset (gas holdup, bubble size distributions and local flow measurements). Unfortunately, the previous study suffers from some limitations; in particular, in the previous experimental dataset, the bubble size distributions concerned only one axial position and a detailed characterization of the gas sparger was missing. This study contributes to the existing discussion and proposed a step ahead in the study of the “pseudo-homogenous” flow regime. To this end, we propose an experimental study, to improve the comprehensive dataset previously obtained. The novel dataset—obtained for two gas velocities—concerns bubble size distributions at different axial and radial positions and a precise characterization of the gas sparger. The comprehensive bubble size distribution dataset may serve as basis to improve the coalescence and break-up closures; conversely, the precise characterization of the gas sparger served as an improved input to the numerical simulations. The numerical results, with two different lift force implementations, have been compared with the whole dataset and have been critically analyzed. Reasons for the discrepancies between the numerical results and the experimental data have been identified and may serve as basis for future studies