Experimental results of Gas-liquid Flow Pattern in Pipe

Published: 7 October 2020| Version 3 | DOI: 10.17632/k637x42yxx.3
André Quintino


The dataset contains experimental data collected in the literature for gas-liquid flow, this data was used in the paper "Flow pattern transition in pipes using data-driven and hybrid-physics-data machine learning". This data is compiled in the excel spreadsheet herein. Where the column A and B indicates the reference in which the data was extracted. The data was visually collected through the flow pattern maps published in those works (column C). Therefore, minor deviations on the superficial velocities of the data may occur if compared with the raw data. Each experimental point in the database contains the experimental apparatus characteristics (duct material, development length, duct internal diameter (d) and inclination of the pipe (θ)), operational pressure, temperature, working fluids and their properties (viscosities (μ_L and μ_G), densities (ρ_L and ρ_G) and interfacial tension (σ)), superficial velocities of each of the phases (U_sL and U_sG) and observed flow patterns.



Universidade de Sao Paulo Campus de Sao Carlos


Fluid Mechanics, Machine Learning, Multiphase Flow, Flow Stability, Hydrodynamic Stability