搅拌反应器内液—液分散特性的CFD-PBM数值模拟

发布时间：2019-04-09 17:05

【摘要】：液-液搅拌反应器被广泛应用于冶金、化工、制药等行业中,是促进相间混合、传质、传热,实现萃取、结晶、分散等过程的重要工业设备。计算流体力学(CFD)在用于研究液-液分散体系的过程中,由于相间耦合及液滴破碎和聚并的复杂性,现有模型很难对其进行精准的预测。为此,本文结合CFD和群体平衡模型来对搅拌反应器内的液-液分散体系进行深入研究,研究结果可为大型液-液搅拌反应器的优化设计提供参考和指导。本文对CFD中的可压缩双流体模型进行详细分析,结合群体平衡模型求解方法中的积分矩方法(QMOM)和拓展积分矩方法(EQMOM),对控制方程中的压力泊松方程、矩传输方程等进行重构优化,实现了并行求解稳定性、相分数的有界性及守恒性、矩的有界性及可实现性,同时将CFD与群体平衡模型耦合并植入到OpenFOAM中,模拟结果和实验数据吻合。表面活性剂会直接影响液-液分散体系内的液滴破碎和聚并行为。本文选用QMOM求解群体平衡模型并结合CFD对附加表面活性剂的液-液分散体系进行研究。研究结果表明表面活性剂使得液滴聚并行为难以发生,MB和Lehr破碎模型的预测平均Sauter直径和实验数据偏离较大,改进后的MF破碎模型预测的平均Sauter直径可以和实验数据较好的吻合且优于CT和Alopaeus破碎模型的预测结果。对于低离散相相含率的液-液分散体系,由于液滴碰撞的可能性较小以致聚并行为很难发生。本文从文献中选取不同的实验数据,使用不同的破碎模型研究离散相粘度及放大效应对液滴破碎的影响。CT破碎模型假定液滴的破碎频率主要取决于湍流动能耗散率。MF破碎模型认为液滴的破碎频率和湍流动能耗散率、湍流动能、湍流尺度以及其他的流体属性均相关。研究结果表明CT破碎模型不能够捕获湍流间歇性以及离散相粘度引起的液滴平均Sauter直径变化,MF破碎模型可以很好的捕获这种平均Sauter直径的差异。当离散相相含率增加时,液滴之间的破碎以及聚并过程同时存在。离散相相含率对液滴聚并过程的影响可通过“湍流阻尼因子”来实现。本文选取文献中不同的中等离散相相含率的液-液分散体系,结合CT和MF模型对液滴的破碎及聚并过程进行研究。研究结果表明基于连续数量密度函数的log-normal EQMOM优于常规的QMOM方法,CT模型预测的破碎和聚并过程较为迅速,MF模型预测的破碎和聚并过程较为动态且更好的反映了平均Sauter直径实验数据的动态变化过程。湍流阻尼因子的引入也有助于使得模拟结果更加准确。
[Abstract]:Liquid-liquid stirred reactor is widely used in metallurgy, chemical industry, pharmaceutical industry and so on. It is an important industrial equipment to promote interphase mixing, mass transfer, heat transfer, extraction, crystallization, dispersion and so on. The computational fluid dynamics (CFD) model is difficult to accurately predict the liquid-liquid dispersion system due to the coupling between phases and the complexity of droplet fragmentation and coalescence in the study of liquid-liquid dispersion system. Based on CFD and population equilibrium model, the liquid-liquid dispersion system in stirred reactor is studied deeply in this paper. The results can provide reference and guidance for the optimization design of large-scale liquid-liquid stirred reactor. In this paper, the compressible two-fluid model in CFD is analyzed in detail. Combined with the integral moment method (QMOM) and the extended integral moment method (EQMOM), the pressure Poisson equation in the governing equation is combined with the integral moment method (QMOM) and the extended integral moment method (EQMOM),). The moment transfer equation is reconstructed and optimized to realize the stability of parallel solution, the boundedness and conservation of phase fraction, the boundedness and realizability of moment. At the same time, the CFD is coupled with the population equilibrium model and implanted into OpenFOAM. The simulation results are in good agreement with the experimental data. Surfactant has a direct effect on droplet fragmentation and coalescence in liquid-liquid dispersion system. In this paper, QMOM is used to solve the population equilibrium model and CFD is used to study the liquid-liquid dispersion system with additional surfactants. The results show that surfactant makes the coalescence of droplets difficult to occur, and the predicted average Sauter diameter and experimental data of MB and Lehr fragmentation models deviate greatly. The average Sauter diameter predicted by the improved MF fragmentation model is in good agreement with the experimental data and is better than that predicted by the CT and Alopaeus fragmentation models. For liquid-liquid dispersion systems with low discrete phase holdup, the coalescence behavior is difficult to occur due to the low probability of droplet collision. In this paper, different experimental data are selected from the literature. Different fragmentation models are used to study the effects of discrete phase viscosity and amplification effects on droplet fragmentation. Ct fragmentation model assumes that the breakup frequency of droplets is mainly dependent on turbulent kinetic energy dissipation rate. MF fragmentation model considers that the breakup frequency of droplets is mainly dependent on turbulent kinetic energy dissipation rate. Rate and dissipation rate of turbulent kinetic energy, Turbulence kinetic energy, turbulence scale and other fluid properties are all related. The results show that the CT fragmentation model can not capture the turbulent intermittency and the variation of the mean Sauter diameter caused by the viscosity of the discrete phase. The MF fragmentation model can capture the difference of the average Sauter diameter well. When the phase holdup of the discrete phase increases, the fragmentation and coalescence process of the droplets exist at the same time. The effect of discrete phase holdup on droplet coalescence can be realized by "turbulent damping factor". In this paper, different liquid-liquid dispersion systems with medium discrete phase holdup are selected in the literature, and the breakup and coalescence process of droplets are studied by means of CT and MF models. The results show that the log-normal EQMOM based on the continuous quantitative density function is superior to the conventional QMOM method, and the fragmentation and coalescence process predicted by the CT model is faster than that based on the continuous quantitative density function. The fragmentation and coalescence process predicted by the MF model is dynamic and better reflects the dynamic change process of the experimental data of the average Sauter diameter. The introduction of turbulent damping factor also helps to make the simulation results more accurate.
【学位授予单位】：北京化工大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TQ052.5

【参考文献】