Predictive Model for Particle Transport Velocities in Multiphase Gas-Liquid Flows using Artificial Intelligence

The purpose of this work is to investigate the use of several Machine Learning (ML) models to predict the critical velocities of various single-phase and multi-phase carrier fluids in horizontal and inclined flow conditions. A methodology is developed to predict critical velocities in pipes via ML, using accessible parameters as inputs, namely, fluid and particle properties and inclination angles. The ML algorithms are trained on a large dataset (more than 2000 data points) of critical velocities in single and multiphase carrier fluids that are collected from open-source: articles and dissertations. The proposed ML approach is observed to have good performance across a wide range of flow conditions and inclination angles. The final objective of this project is to develop an Artificial Intelligence/ML tool for predicting the critical velocity in multiphase flows.