Gas-liquid test cases simulated by the Eulerian-Eulerian method

[李东岳]

Abstract

The Eulerian-Eulerian (E-E) approach is widely used to simulate gas-liquid system since it is applicable to a wider range of phase fraction. However, to obtain satisfactory predictions, suitable sub-models and parameters need to be adjusted. In this work, ten classic gas-liquid systems, as the baseline test cases, were selected to validate the E-E approach numerically with fruitful experimental data in the literature. These test cases feature with different disciplines and were selected from different industries covering from chemical, nuclear, bioprocessing, metallurgical and ocean engineering. Simulations were launched by the OpenFOAM solver reactingTwoPhaseEulerFoam, in which an E-E method was implemented with sophisticated numerical techniques. The governing equations and numerical methods were firstly revisted in details. The effects of the mesh resolution, discretization scheme, turbulence model, drag force and non-drag forces on the predicted phase fraction distribution, gas holdup, liquid velocity and the turbulence variables were investigated. Our results show that the including of the drag force, as the most important momentum interfacial exchange term, is very important for all the test cases. It determines the global gas holdup and the vertical liquid velocity. Other non-drag forces play a crucial role for the test cases of which the aspect ratio of the geometry is large (e.g., pipe flows). Meanwhile, relatively coarse mesh can be used for all the test cases since the E-E approach is based on the macroscopic assumptions. The turbulence model and the bubble induced model can improve the predictions in certain cases. However, a universal model setting can be hardly obtained.

Test cases

Results

[金石为开]

@东岳 老师 你好我最近也在研究reactingtwophaseeulerfoam（a）之前研究过twophaseeulerfoam （b）我不研究传热 我看a模型里面有表面张力 而b里面没有 想问下他们在表面张力上为什么一个考虑一个不考虑 对于模型有什么区别呢 谢谢

[李东岳]

@金石为开 reactingTwoPhaseEulerFoam里面添加了相界面压缩的功能，如果关闭的话，表面张力没用

[金石为开]

@东岳 是吗 我分别用reactingtwoPhaseeulerfoam和twophaseeulerfoam算 相同的相间作用力模型 计算的结果不一样 唯一多的是前者有surfacetension 我想solver计算了 我关闭tension就不能计算 感觉不太一样啊

[金石为开]

@东岳 按照老师讲的reactingTEE模型没有表面张力 应该和twophaseeulerfoam计算结果一样 可是计算不太一样 可能是什么原因呢

[李东岳]

@金石为开 把你的数据贴上来看看

[金石为开]

横坐标通气速度 纵坐标减阻率 两个曲线不一样 我算的是一个二维的平板减阻 两个solver模型一样 曲线后面差异很大

[金石为开]

@东岳 twophaseEulerfoam和reactingtwophaseEulerfoam看了看 代码写的很不一样 后者solver代码有点难看懂 尤其是模板那部分 还有怎么调用各种相间作用模型的代码不好看懂 我一直看的前者solver 最近才看reactingtwophaseEulerfoam 主要是想用用的研究一下它的polydisperse群体平衡模型 虽然目前小白一个

[李东岳]

正如你所说，这俩个求解器算法有很大变化，你这个表现得很明显。如果有空的话，你可以给基金会提bug，目前涡暂时没有时间详细对比两个求解器。

针对你的问题，我更倾向于跟实验对比，哪个更准用哪个。我模拟的最后一个算例也是气泡减阻。看主楼的Case D.1。参考的是Mohanarangam2009的文章

[金石为开]

老师 我刚刚比较算了下 op6的reactingtwophaseeulerfoam和op300的reactingtwophaseeulerfoam计算的不一样 op300的reactingtwophaseeulerfoam的计算和两个版本twophaseeulerfoam计算一样 看来是op6版本的reactingtwophaseeulerfoam改变了 不知道里面有什么算法之类的改变吧

[李东岳]

我觉得你应该跟实验数据比一下

[李东岳]

@金石为开 Any updates?