摘要

The distribution of downward air-water flow in vertical 32.5 mm feeder tubes connected to a horizontal 203 mm cylindrical header was studied experimentally in a modular header facility operating at near-atmospheric pressure. Several experimental configurations with different combinations of two or four feeders were used. A large database of mass flow rate measurements was compiled using a flow rate measurement method based on machine learning. The collected data allowed the effects of inlet flow rates and feeder location on the gas and liquid flow distribution to be studied. The feeder mass flow rates were also qualitatively correlated to the observed flow patterns in the header, including the inlet gas-liquid jet, the free surface level and the cross-flow velocity over the feeders. Three potential approaches for correlating the wall pressures measured at different locations to the flow rate distribution were tested. Although under certain conditions a measurable pressure was found to be correlated with the flow distribution, it was concluded that none of these methods was capable of estimating consistently and accurately the flow distribution in a header-feeder system.

  • 出版日期2015-9