The maldistribution in the parallel multiple channels commonly leads to a reduction in the efficiency of heat exchangers and even causes the equipment to suffer the risk of failure. The distribution of gas–liquid two-phase slug flow when moving through the parallel multiple channels is still an unresolved problem and limited by the lack of experimental data. This work presents an experimental study on the air–water two-phase flow in the parallel multiple channels characterized by a horizontal header and 2~4 vertically-upward branches under the slug flow pattern conditions and develops a new predictive model. The effects of inlet superficial velocities and branch number have been investigated in the ranges of inlet gas and liquid superficial velocities of 1.4~25.0 and 0.2~1.7 m·s⁻¹, respectively. Results show that the distribution characteristics of two-phase slug flow highly depend on the inlet superficial velocities which manifest in that the peak value of liquid phase flow ratio shifts to the downstream branches sequentially as the inlet superficial velocities increase. Decreasing the branch number is beneficial for the uniform distribution of each phase. A predictive model is proposed for the distribution of air–water slug flow by considering the effects of superficial velocities and branch number. The new model achieves the agreement of 91.3% for prediction distribution within the range of ±25% when compares to present experimental data and shows good agreement with previously reported experimental results of both T-junction and parallel multiple channels. The acquired experimental data and predictive model are crucial for the development and optimal design of the heat exchangers.

并行多通道中的分布不均通常会导致热交换器效率降低，甚至导致设备遭受故障的风险。气液两相段塞流经平行多通道流动时的分布仍然是一个尚未解决的问题，并且由于缺乏实验数据而受到限制。这项工作提出了在团状流型条件下以水平集管和2〜4个垂直向上分支为特征的平行多通道中气水两相流的实验研究，并开发了一种新的预测模型。在1.4〜25.0和0.2〜1.7 m·s ^-1的范围内，研究了进气表面速度和支路数的影响。， 分别。结果表明，两相段塞流的分布特征在很大程度上取决于入口的表观速度，表现为随着入口表观速度的增加，液相流率的峰值依次向下游分支移动。减少分支数有利于各相的均匀分布。通过考虑表层速度和分支数的影响，提出了一种空气-水弹团流量分布的预测模型。与目前的实验数据相比，新模型在±25％的范围内的预测分布达到了91.3％的一致性，并且与T结和平行多通道的先前报道的实验结果显示出良好的一致性。