流态化模拟:基于介尺度结构的多尺度CFD

介尺度结构是研究气固流态化多尺度行为的关键。传统的基于平均化处理方式的双流体模拟不能准确描述流化床中的多尺度流动和传递行为。相较而言,基于能量最小多尺度(EMMS)方法的结构多流体模型(SFM)基于局部空间(网格)内的非均匀介尺度结构流动特征,其宏观预测结果与网格分辨率基本无关,因而可以大幅降低模拟计算量。基于SFM模拟得到的流动结构,EMMS多尺度传质模型进一步成功解释了传统传质文献中的数据差异。集成上述模型,形成了一整套模拟流化床流动-传递-反应耦合过程的多尺度计算流体力学(CFD)方法,并将其应用于预测循环流化床中典型的S型轴向分布、揭示噎塞转变的机理以及流化床放大困难的原因。多尺度CFD使工业规模循环床的三维、全系统、动态流动-反应耦合过程的准确模拟成为可能,并为实现从模拟向实时虚拟过程转变的目标打下基础。

Fluidized bed simulation: structure-dependent multiscale CFD

Meso-scale structure is key to gas-solids fluidization modeling.Traditional two-fluid model is not suitable for describing multiscale behavior in fluidized beds.In contrast, the structure-dependent multi-fluid model (SFM), which is based on the energy-minimization multi-scale (EMMS) method, takes into account the heterogeneous structure within local space (or, sub-grid structure) and hence its prediction is only weakly dependent on grid resolution, reducing greatly computing load.Based on the structural characterization of SFM, we proposed an EMMS based mass transfer model to explain why literature data scatter up to several orders of magnitude.These models were integrated into a whole set of multiscale computational fluid dynamics (CFD) method, with which we predicted the typical S-shaped axial profile of volume fraction and further revealed the "choking" mechanism and the reason why scale-up of a fluidized bed was so difficult.Currently, 3D, whole-loop, transient, reactive simulation of an industrial circulating fluidized bed could be made possible.In prospect, to realize the shift of research mode from simulation to virtual process engineering, we still need breakthrough in understanding meso-scale structures.