湍流场中液滴群破碎行为研究一直是液液两相流领域一个重要问题，其破碎机理的揭示和理论模型的建立对深化液液两相流基本规律认识、实现两相流行为准确预测、提高两相流设备设计和运行水平有重要意义。本论文以揭示湍流流场中液滴破碎机理为目标，对液滴群破碎行为开展了系统的实验和理论研究。利用高速摄像技术实现了搅拌槽内液滴群破碎核函数的直接测定，系统考察了转速、液滴尺寸、界面张力和分散相粘度等主要因素对液滴群破碎行为的影响规律。发现在稳态操作条件下液滴以二元破碎为主，液滴破碎模式可划分为初始拉伸破碎、中间拉伸破碎和旋转式破碎三类，对应的子液滴尺寸分布分别为倒U形、U形和M形。通过引入湍动应力、界面应力和液滴粘性应力构建了破碎频率和子液滴尺寸分布的经验模型，模型的预测结果与实验数据符合较好。基于对液滴破碎过程和破碎时间实验数据的定量分析，提出了新型的液滴振荡破碎机理。基于此机理，利用量级分析和特征参数法构建了湍流条件下液滴破碎时间的理论模型，模型预测和实验结果均表明在分散相粘度和湍动强度不太高的条件下破碎时间正比于液滴的二阶振荡周期，进一步证明了振荡破碎机理的合理性。基于液滴振荡速度的三维Maxwell分布构建了液滴破碎概率的理论模型，结果显示破碎概率与Weber数和Ohnesorge数密切相关。结合上述两个模型构建了液滴破碎频率的理论模型，并利用实验数据验证了模型的准确性和普适性。利用所构建的液滴破碎核函数模型，通过耦合计算流体力学和群体平衡模型，开展了泵轮式混合槽内液-液分散特性的数值模拟。模拟结果表明混合槽内存在两个纵向循环区和一个相对静止区，液滴破碎主要发生在叶片附近的高湍动能耗散区域，而液滴聚并发生在整个混合槽内。此外，分散相存留分数和液滴尺寸空间分布的均匀性均随着循环强度的增加而增加。利用上述模拟方法，实现了中试规模泵轮式混合澄清槽液液两相流的数值模拟，将模拟得到的液滴尺寸分布应用到了混合澄清槽传质实验数据的分析，结合传质模型得到了Sherwood数，为相关设备设计放大提供了依据。通过与基于液滴Sauter平均直径计算的Sherwood数的对比，也证明了利用群体平衡模型计算液滴尺寸分布的必要性，它使Sherwood数有更好的规律性，传质模型具有更准确的预测能力。

The study on the breakup behavior of drop swarms in turbulent flow is one of the critical subjects in liquid-liquid two-phase flow. The revealing of the breakup mechanism and the establishment of the theoretical model are of great significance for deepening the understanding of the fundamental law of liquid-liquid two-phase flow, realizing the accurate prediction of flow behaviors, and improving the design and operation level of the two-phase flow equipment. To reveal the drop breakup mechanism in turbulent flow, in this dissertation, systematic experimental and theoretical studies on drop breakup were conducted.Kernel functions describing the drop breakup in a stirred tank were directly measured by the high-speed camera technology. Influences of the rotating speed, drop diameter, interfacial tension, and drop viscosity on drop breakup were quantitatively investigated. The binary breakup was dominated under the steady-state operation condition. Moreover, three breakup patterns, i.e., original tensile breakup, intermediate tensile breakup, and revolving breakup were observed, daughter drop size distributions (DDSD) of the three breakup patterns were inverted U-shaped, U-shaped, and M-shaped respectively. By introducing the turbulent stress, interfacial stress, and viscous stress of the drop, empirical correlations of drop breakup frequency and DDSD were proposed and good agreement was found between the prediction results and the experimental data.Based on the quantitative analysis of experimental data on the drop breakup process and breakup time, a novel mechanism of drop oscillation breakup was proposed. Based on this mechanism, a theoretical model of drop breakup time under turbulent flow was constructed by using order-of-magnitude analysis and the characteristic parameter method. Both model predictions and experimental results show that the breakup time is proportional to the second-order oscillation period of the drop under conditions where the drop viscosity and turbulence intensity are not too high, which further validates the oscillation breakup mechanism. Furthermore, a theoretical model of drop breakup probability was established based on the three-dimensional Maxwell distribution of drop oscillation velocity, results show that the breakup probability is determined by Weber number and Ohnesorge number. Combining the two models above, the theoretical model of drop breakup frequency was obtained, the accuracy and generality of the model were then validated by comparing with experimental data.Using the constructed breakup models, numerical simulations of the liquid-liquid dispersions in a pump-mixer were performed by coupling the computational fluid dynamics and population balance model. Flow structures show that two recirculation loops and one quiescent region exist in the pump-mixer. Breakup regions were mainly located near the impeller where high energy dissipation rates exist, while the coalescence regions were observed in the whole mixer. Moreover, both the holdup of the dispersed phase and the Sauter mean diameter were uniformly distributed in space at higher flow recirculation intensity.Using the above simulation method, the numerical simulation of liquid-liquid two-phase flow in a pilot-scale mixer settler was performed, and drop size distributions under different operating conditions were obtained. Furthermore, the Sherwood number was obtained by combining the experimental data and mass transfer models, which provides a basis for the design and scaling of the relevant equipment. By comparing with the Sherwood number calculated based on the Sauter mean diameter, the necessity of using the population balance model to calculate the drop size distribution is also demonstrated, which makes the Sherwood number have better regularity and the mass transfer model have more accurate prediction ability.