Estimation of gas hold-up in bubble columns using wall pressure fluctuations and machine learning

In collaboration with the University of Limerick and Helmholtz Zentrum Dresden Rossendorf, we have developed a novel method to estimate gas hold-up in bubble columns using wall pressure fluctuations. This machine learning-based approach leverages an artificial neural network (ANN) to predict gas hold-up accurately across varying operating conditions, column configurations and physical properties. The study demonstrates the reliability of this method even with unseen experimental data, marking a significant advancement in non-invasive characterization of multiphase flow systems.

Estimation of gas hold-up in bubble columns using wall pressure fluctuations and machine learning