中文摘要：

在湍动流化床中，过渡段对于包括甲醇制烯烃在内的气固催化快反应有着重要的作用。采用PV6D反射型光纤探针对内径95mm的湍动流化床内过渡段的固含率分布和脉动参数进行了测量，分别考察了表观气速和静床高的影响，并采用修正的基于颗粒动力学的三段曳力双流体模型进行模拟。实验表明，湍动流化床过渡段中固含率的轴向分布呈现S型和指数型两种类型，固含率轴向与径向分布都在过渡段内出现最大梯度，表明过渡段中固体浓度分布比稀相段和密相段更不均匀。表观气速和静床高的变化将导致S型和指数型分布的相互转变，并且对过渡段底部与壁面附近的固体高浓度区影响最为显著。局部固含率脉动概率密度分布表明，在静床高较小时，随着气速的增大，床层下部气含率最大值位置将从中心区移动至环隙区，呈现气含率的双峰型分布。本文提出的修正三段曳力模型考虑了颗粒团聚的影响，对过渡段中分布板影响区之外的固含率分布均能较好地模拟。

英文摘要：

The transition section between bottom dense zone and top dilute zone is important for fast gas- solids catalytic reactions, such as methanol to olefins （MTO）. Solids concentration and fluctuation parameters in the transition section of a fluidized bed with 95 mm i. d. were measured by using reflective optical fiber probe （PV6D）. The influences of superficial gas velocity and static bed height were investigated. An Eulerian-Eulerian two-fluid model that incorporated kinetic theory of granular flow and modified three-zone drag law was used to predict the flow structure in the turbulent fluidized bed. The axial and radial distributions of solids concentration were simulated by means of computational fluid dynamics （CFD）. The experiments showed two axial profiles of solids distribution in the transition section~ exponential and S-shaped. The maximum gradients of both axial and radial solids concentration profiles were located in the transition section, which meant that solids distribution in the transition section was much more non-uniform than that in the dilute and dense sections. The change of gas velocity and static bed height would cause a shift between exponential and S-shaped profiles, and significantly affect solidsconcentration in the dense region near the wall and at the bottom. Probability density distribution of local solids concentration revealed that when static bed height was small, the location of maximum voidage would gradually move from core region to annular region with increasing gas velocity, resulting in a bimodal distribution of voidage. A modified three-zone drag model was proposed by taking the effect of particle clustering into account for CFD simulation. The simulated solids concentration distribution agreed well with experimental data in the transition section except for the distributor zone.