Two-phase flows in
pipes develop characteristic patterns or flow
morphologies. Their transport between process units and plant components
often involves short pipes, large pipe diameters, as well as bends
and curvatures. The prediction of the predominantly undeveloped flow
morphologies in such systems is challenging and subject to high uncertainties
due to lacking experimental data and universal engineering models.
In this work, comprehensive experimental studies were conducted in
horizontal straight and bent pipes of 50 mm and 200 mm diameter. Wire-mesh
sensors were applied at characteristic positions to obtain the gas–liquid
distributions with high spatiotemporal resolution. Subsequently, a
fuzzy classification method is applied to assign a flow pattern to
characterize the flow conditions. As this assignment is fuzzy, we
introduce an advanced concept of a color-coded flow map visualization
for further analyses. As a result, we analyze the effect of pipe curvatures
on the flow morphology and its downstream recovery.