Knowledge Management System Of Institute of process engineering,CAS
气固流化系统是一个典型的非线性非平衡系统，呈现出复杂的多尺度特性：如，局域空间颗粒浓度的非均匀分布和颗粒速度的非高斯分布、时空交替的介尺度结构以及床层整体随表观气速而变化的流域等。这些复杂特性与单个颗粒之间的非弹性碰撞和摩擦、气体与颗粒以及颗粒群之间的相互作用、气固两相湍流等因素紧密相关，是流化床模拟计算的核心难题，也是突破工业反应器放大、设计和优化的关键。为此，有必要从单颗粒层次的运动行为出发，对流化床的多尺度非平衡特性展开系统深入的研究。针对以上目标，本论文结合实验和计算方法对流化床中局域非平衡特性、介尺度结构特性以及表观气速对颗粒运动状态的影响等展开研究。本文主要内容和结果如下：1. 应用高速摄像机拍摄床层中颗粒运动，通过颗粒跟踪测速法（PTV）和Voronoi划分方法得到单颗粒的速度和空隙率。统计鼓泡床和湍动床中颗粒速度概率密度分布、平均空隙率、颗粒平均速度、颗粒温度、颗粒湍动能等物理量，发现浓相和稀相中颗粒的瞬时颗粒速度概率密度分布接近高斯分布，各物理量尺度依赖性和各向异性较弱，呈现近局域平衡特性。而两相界面处颗粒速度概率密度分布严重偏离高斯分布，甚至会出现双峰分布，各固相宏观物理量尺度依赖性和各向异性较强。这些结果揭示了气固流化床中的局域非平衡性，表明尺度分离假设难以成立，因此传统的双流体模型和颗粒动理论难以对该体系进行准确描述。2. 使用高解析度的实验数据对介尺度结构展开动力学分析。考察鼓泡床中气泡直径和气泡运动速度，发现本实验中气泡运动规律可以使用经典模型对其进行描述。根据颗粒Voronoi分布提出了一种判定团聚物的方法，对湍动床中的团聚物平均速度、平均颗粒温度以及团聚物的尺寸分布等时均特性进行分析，深入了解湍动床中团聚物的运动状态。同时对团聚物的动态变化过程进行细致考察，分析团聚物聚并和破碎过程中相应颗粒性质的变化，发现团聚物聚并过程可以使用扫雪机模型描述，在本实验中颗粒动能的损失和颗粒所占面积减少量都与t3/2成正比；而破碎过程中向上和向下运动颗粒动能增长趋势完全不同，表明该过程中颗粒受力作用复杂。3. 通过对比实验中不同气速下时均统计结果，发现气速增加会导致时均颗粒脉动速度分布更加偏离高斯分布。同时，用于描述颗粒脉动的两个物理量——颗粒温度和颗粒湍动能显示出较强的尺度依赖性，二者的加和——颗粒总脉动能在长时间统计下显示出尺度无关性，有助于建立稳态模型。进一步分析颗粒总脉动应力，发现其在鼓泡床中作用很小，但是在湍动床中，其空间梯度分布引起固相在竖直方向受力十分明显，表明在高气速条件下的固相应力建模十分重要。4. 应用计算流体力学-离散元法（CFD-DEM）对实验涉及的不同流域工况条件进行模拟计算。以颗粒平均浓度、平均速度和总脉动能作为对比数据，发现模拟计算可以部分再现鼓泡床的行为，但是湍动床模拟结果与实验差距较大，这印证了高气速下的流化床中局域非平衡性更强的实验结果。因此，未来的固固应力及气固曳力建模都应考虑颗粒运动的局域非平衡特性。综上所述，本论文对流化床中不同尺度的非平衡特性进行了深入的研究。从局部颗粒速度分布、固相宏观物理量的尺度依赖性以及各向异性等个方面揭示了流化床中的局域非平衡性；对床层中介尺度结构进行研究，提出了团聚物的判定方法，统计并分析了团聚物的时均和动态特性；通过考察表观气速对局域非平衡性的影响，进一步强调宏尺度条件对于流态化建模的重要性。这些结果有助于揭示流态化复杂多尺度特征背后的机制，从而为气固两相流物理建模以及流化床模拟计算、放大与设计提供基础。;Gas–solid fluidization is a typical nonlinear nonequilibrium system with multiscale nonequilibrium features in terms of locally heterogeneous distribution of particles and non-Gaussian distribution of particle velocity, spatio-temporal evolutions of the meso-scale structures, and regime transitions with superficial gas velocity. These features, closely related to the inelastic collision and friction among particles, the interaction between gas and particles and particle groups, the gas-solid two-phase turbulence and other factors, are the core challenges of simulations and play significant roles in industrial reactor scale-up, design, and optimization. Therefore, it is necessary to study multiscale nonequilibrium features from the microscopic statistics.To this end, we studied the local nonequilibrium features, characteristics of mesoscale structures and effects of superficial gas velocity on fluidization behavior through experimental and numerical methods. The main contents and results of this thesis are as follows:1. A high-speed camera was used to capture the particle motion, and then the velocity and void fraction of individual particle were obtained by using particle tracking velocimetry (PTV) and Voronoi tessellation methods, respectively. Statistical properties of particles were investigated based on the experimental results of the bubbling and turbulent beds, including the probability density distribution of particle velocity, averaged void fraction, averaged particle velocity, granular temperature, particle turbulent kinetic energy and other physical quantities. It was found that the probability density distribution of particle velocity was close to the Gaussian distribution, and the scale dependence and anisotropy of macroscopic variables were weak in the dense phase and the dilute phase. However, the probability density distribution of particle velocity near the interface between the dense and dilute phases deviated from the Gaussian distribution, and even showed a bimodal distribution. The corresponding physical variables showed strong scale dependence and anisotropy. These results revealed that the gas-solid fluidized bed was local nonequilibrium, especially lack of scale separation. In other words, the hypotheses of the traditional two-fluid models and kinetic theory of granular flow were invalid.2. Kinetic analysis of mesoscale structures were made through highly resolved experimental data. By examining the bubble diameter and bubble rise velocity in the bubbling bed, we found that the features of bubbles could be described with classic models. A new method to identify clusters was proposed based on the Voronoi distribution of particles, and then the time-averaged properties of clusters, e.g., time-averaged velocity and granular temperature of clusters and cluster size distribution, in the turbulent bed were analyzed to gain deeper insight about the flow patterns of clusters. Moreover, dynamic behaviors of clusters, particularly the cluster coalescence and breakup, were examined by analyzing the properties of particles taking part in these processes. It was found that the snowplow model could describe the cluster coalescence process during which the loss of particle kinetic energy and the reduction of the total particle area were proportional to t3/2. Nevertheless, growth tendencies of kinetic energy of upward and downward particles were completely different when clusters breaking, illustrating complex forces on these particles.3. Comparing time-averaged statistical results with different gas velocities, we found the time-averaged distribution of particle fluctuation velocity deviates much from the Gaussian distribution at higher gas velocities. Meanwhile, the granular temperature and particle turbulent kinetic energy showed strong scale dependence, but the sum of them, i.e., the total particle fluctuation energy, seemed more suitable for the steady-state modelling due to its weak scale dependence. Further analysis of the total particle fluctuation stress revealed its minor role in the bubbling bed. However, vertical forces induced by the total particle fluctuation stress showed a remarkable role in the turbulent bed, demonstrating the importance of solid stress modelling for fluidization with high gas velocity.4. The averaged solid volume fraction, particle velocity and total fluctuation energy at various gas velocities were simulated through CFD-DEM and compared with experimental data. It was found that the simulations partially described the behavior of the bubbling bed, but the results of the turbulent bed were poor. These results confirmed the conclusion mentioned in experimental analyses that the local nonequilibrium was stronger for cases with higher gas velocities. It also demonstrated that we needed to consider the local nonequilibrium characteristics for the construction of not only solid-solid stress models but also gas-solid drag models.In summary, multiscale nonequilibrium features were thoroughly studied in this thesis. The local nonequilibrium of gas-solid fluidization was uncovered through analyses of local distribution of particle velocity, scale dependence and anisotropy of macroscopic variables. A new method for cluster identification was proposed, and both the time-average and dynamic characteristics of clusters were statistically analyzed. Further, by studying the effects of superficial gas velocity on the local nonequilibrium, the importance of gas velocity was stressed in the modelling of gas-solid fluidization. All these results would help us understand the complex mechanisms underlying the multiscale characteristics and pave a substantial foundation of modelling and simulations of gas-solid fluidization.
|王海峰. 气固流态化的多尺度非平衡特性研究[D]. 中国科学院大学,2020.|
|Files in This Item:|
|气固流态化的多尺度非平衡特性研究.pdf（11706KB）||学位论文||限制开放||CC BY-NC-SA||Application Full Text|
|Recommend this item|
|Export to Endnote|
|Similar articles in Google Scholar|
|Similar articles in Baidu academic|
|Similar articles in Bing Scholar|
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.