多晶硅流化床反应器内气固两相流场与气泡尺寸分布的TFM-KTGF模拟

硅烷流化床法是生产太阳能级多晶硅的重要技术,流化床反应器床层内的气泡大小和分布特性是影响晶体硅纯度和致密度的关键因素。使用TFM-KTGF两相流模型,基于Ansys Fluent软件平台模拟计算了多晶硅流化床内气固两相流场。在200倍粒径的网格尺寸条件下,对直径0.5 m的圆柱形流化床多晶硅反应器内等温流场进行了模拟,借助Matlab图像处理工具对气含率进行后处理,得到多晶硅流化床内气泡尺寸分布。比较了经典的Gidaspow曳力模型和文献报道的亚网格修正曳力模型(SGS)对气泡尺寸分布的影响。结果表明:由SGS模型得到的平均气泡尺寸沿床高变化规律与Mori-Wen经验公式较吻合,最大偏差是12.6%,小于Gidaspow模型的21.4%;使用不同的曳力模型对气泡分布特性有较大影响。这些结果为进一步开展多晶硅反应器的热态模拟研究提供了基础。

TFM-KTGF Simulation of Gas-Solid Fluid Dynamics and Bubble Size Distribution in A Silicon CVD Fluidized Bed Reactor

Fluidized bed reactor for silane deposition is an important technology for solar-grade silicon production. The gas-solid fluid dynamics and gas bubble size distribution are key factors influencing the purification and density of the crystalline particles. In this paper, the TFM-KTGF two-fluid model coded in Ansys fluent platform was employed in the simulation of the gas-solid fluid dynamics in a silicon CVD(Chemical Vapor Deposition) fluidized bed reactor. A 0.5 m diameter cylindrical silicon fluidized bed was simulated with computational grids of 200 times particle size. The Matlab image processing toolbox was used for the post-processing of the simulation data and to obtain the bubble size distributions. The bubble size distributions calculated by classical Gidaspow drag force model and a literature reported sub grid scale model(SGS) were compared. The results show that the change of the average bubble size distribution along the bed height predicted by SGS model is close to that by Mori-Wen's equation, with a maximum deviation of bubble size of 12.6%, which is smaller than 21.4% obtained from Gidaspow model. The type of drag force model showed relatively strong impact on the distribution of babble size. These results paved a solid foundation for further quantitative analysis of hot model silicon-CVD reactors.