Development of a one-dimensional gas-liquid two-phase flow model for whole flow regime in a vertical pipe

尽管时空演化是垂直管内气液两相最本质的特性之一，但目前还没有可用的数值方法来预测空间演化和流型转变。详细的数值方法，如界面跟踪法，可以预测对空间演化起主导作用的详细的相分布。然而，这些方法需要对气泡合并、聚集等空间演化过程进行建模或假设，并且由于巨大的CPU时间，这些方法在工业设计中的应用是不现实的。本研究的目的是建立垂直管内整个流态的一维气液两相流动模型。为了弥补空间演化数据库的不足，我们在竖直管道中进行了两相流动实验，研究了…引起的体积平均含气率、空泡起伏和流型演变的空间演化规律通过实验结果，进一步建立了气泡聚集和聚结的空间演化唯象模型，并用这些模型进行了一维单向气泡跟踪模拟。实验证实：(1)流型的空间演化主要由气泡分布而不是平均含气率决定，(2)气泡聚集在演化中起主导作用，(3)剪切诱导湍流的存在阻碍了气泡的聚结和聚集，但随着含气率的增加，气泡的聚结和聚集被加强，(4)气泡团的上升速度与相同尺寸的Taylor气泡的上升速度基本相同；(5)大气泡尾迹对小气泡运动的影响随着液体流量的减小而增大；(6)气泡团和Taylor气泡的生长导致空穴沿流动方向的起伏增大。所开发的模拟方法包括气泡的径向分布模型，模拟时间与实际现象基本相同。模拟的流型演变、平均含气率和含气率波动与静止液体、层流和湍流两相流的实验结果吻合较好。较少

Though a space-time evolution is one of the most intrinsic natures of gas-liquid two-phase in a vertical pipe, there is little available numerical method for the prediction of the spatial evolution and the transition of flow regime. Detailed numerical methods such as an interface tracking method can predict detailed phase distributions, which play a dominant role on spatial evolution. These method however need models or assumptions on processes of spatial evolution such as bubble coalescence and clustering, and the application of these method to industrial design is not realistic due to the huge CPU time. The purpose of the present study is to develop a one-dimensional gas-liquid two-phase flow model for whole flow regime in a vertical pipe. In order to make up for the lack of database on spatial evolution, we carried out two-phase flow experiments in a vertical pipe, and investigated spatial evolution of volume-averaged void fraction, void fluctuation and flow pattern evolution caused … More by the bubble clustering and coalescence Phenomenological models for the spatial evolution were developed through the experimental results, and one-dimensional one-way bubble tracking simulation with these models was developed and evaluated.It was confirmed through the experiments that (1) the spatial evolution of flow pattern is dominated by the bubble distributions rather than the mean void fraction, (2) bubble clustering plays a dominant role in the evolution, (3) bubble coalescence and bubble clustering are hindered by the presence of shear-induced turbulence, but they are enhanced with increasing the void fraction, (4) the rising velocity of a bubble cluster is almost equal to that of a Taylor bubble in the same size, (5) the impact of wake behind large bubble on the motion of small bubbles increases with decreasing the liquid flow rate, and (6) the void fluctuation increases along the flow direction due to the growth of bubble clusters and Taylor bubbles.Phenomenological models were developed based on the experimental results, and applied them to one-dimensional bubble tracking method. The developed simulation method includes a model of radial distribution of bubbles, and the time for the simulation is almost the same with the actual phenomena. The simulated flow pattern evolution, averaged void fraction and void fluctuation were agree well with the experimental results of two-phase flow in stagnant liquid, laminar and turbulent liquid flow. Less