悬浮床加氢反应器壁面的结焦预测

Prediction of the Coking on the Wall of the Suspended Bed Reactor for Hydrogenation Reaction

以1个空桶结构的悬浮床加氢反应器为研究对象,首先采用计算流体力学模拟软件来模拟其内部流动场,然后通过MATLAB中的BP神经网络模型以及数据拟合软件对实验数据进行拟合,得到结焦率与反应条件之间的关系式。最后将得到的关系式作为变量插入到计算流体力学后处理软件中进行分析,观察各个壁面处的coke值分布图,并结合内部速度场分布判断易结焦位置。结果表明,在此模型中易结焦位置集中在反应器底部中间位置、分配盘上表面中心区域以及分配盘与反应器主体壁面的连接处。

Choosing a suspended bed hydrogenation reactor with a simple structure as an object, the computational fluid dynamics software was used to simulate the internal fluid flow in it. The BP neural network model in the MATLAB and the data fitting software were used to calculate the experiment data, and finally got the relation formula between the coke yield and the reaction data. The relation formula was used as a variable to insert into the simulation result files of CFD-post. Finally the contour of the coking data can be seen directly from the wall, combining with the internal flow field, the easy-coking position can be estimated. The result shows that, in this model, the easy-coking positions concentrated on the middle on the bottom, the central region of the upper surface on the distribution tray and joint of the tray and the wall.