基于凝结旋流运动映射的超音速分离器多维性能优化研究

Supersonic separator with many excellent characteristics, such as simple and reliable structure, and no adding solvent, is a novel condensing separation equipment for multicomponent gas mixtures and is beneficial for economy and environmental protection. Aiming at the characteristics of the vapor condensation and droplet motion in strong swirling flows of supersonic separator, the flow visualization method based on sensor-CFD data fusion procedure will be proposed. In spatial and temporal scales, the swirling flow motion mapping of condensation droplets and multidimensional correlation model for separation performance can be established. Finally, the multi-objective optimization of the separator performance in feasible region will be accomplished based on experimental data of moist air and natural gas. The highlights are as follows: firstly, based on the characteristics of ‘cryogenic condensation’ and ‘cyclonic separation’ processes of supersonic separator, a data fusion algorithm combining with experiments data of different types of sensors, such as Phase Doppler Particle Analyzer and Laser Particle Analyzer, and CFD results will be developed for completely reconstructing the two-phase flow field. The behaviors of the nucleation, growth, slip, collision and coalescence of the droplets can be extracted, and then the swirling flow motion mapping of condensation droplets will be built. Secondly, combining with the priority ranking and regression methods, the correlation model among structure, fluid and performance will be established. Subsequently, a constrained multi-variable multi-separator optimization method based on Monte Carlo method will be proposed.

超音速分离器是一种新型多组分气体冷凝分离设备，具有简单可靠、结构紧凑、无需添加溶剂等特点，可实现节约成本和保护环境的双重效益。本申请针对超音速分离器强旋流场液滴凝结与运动特征，提出融合多源异构传感器与CFD信息的流场显示方法，在空间和时间尺度上建立凝结旋流运动映射及性能多维度关联模型，并依据湿空气和天然气实流测试数据，实现分离器可行工作域约束下性能多目标优化。其创新在于：(1)依据分离器“低温凝结”和“旋流分离”特征，构建基于相位多普勒分析仪和激光粒度仪的多源异构传感器以及CFD两相凝结双流体模型，并提出异类信息融合算法，形成完整流场显示方法，再提取液滴成核、生长、滑移、碰撞与聚并等过程特征，建立凝结旋流运动映射;(2)以优先级排序法、回归模型法等为基础，建立结构-流体-性能交互关联函数模型，进而提出基于蒙特卡洛法的多变量、多目标和多约束条件的分离器优化方法。

本项目以超音速分离器为研究对象，为解决气液两相旋流场内“低温凝结”和“旋流分离”复杂耦合问题，实现凝结和分离性能优化目标，以数学建模、数值仿真和实验测量三者相结合为手段，对超音速流场内液滴成核、生长、滑移、碰撞与聚并等动力学行为展开了深入研究，揭示了气液两相凝结旋流运动映射机理，并提出了掺混式非均质凝结强化下的性能优化策略。首先，构建了融合多源异构传感器与CFD信息的气液两相流场显示方法，开发了均质/非均质凝结超音速旋流分离与雾化混合实验平台及液滴和液膜测量系统，并分别从欧拉法和拉格朗日法两个角度建立了考虑熵输运方程和示踪剂扩散方程的超音速均质/非均质凝结多分散性液滴CFD模型，实现了超音速分离器内气液两相流与热质传输过程可视化。其次，研究了凝结液滴热力学与动力学特性，获得了凝结Wilson点位置规律、超音速分离器喷雾混合液滴SMD特征和不同粒径与浓度非均质液滴动力学行为，同时基于QMOM方法预测了超音速凝结液滴多分散特性，基于熵输运方程评估了四类熵产和㶲损失分布，并基于液滴与液膜耦合CFD模型，分析了气相、凝结液滴和液膜之间的耦合行为。最后，基于均质/非均质凝结和蒸发双流体模型，研究了不同分离器结构中的性能优化策略，获得了增强分离性能的外加核心最优半径规律，并利用示踪剂扩散方程观测了超音速分离器内液滴掺混传热传质特性，提出了基于掺混式非均质凝结强化的超音速分离器性能优化策略。