气液固流化床气泡特性及气含率预测模型

Bubble Characteristics and Prediction Model of Gas Holdup in the Gas-Liquid-Solid Fluidized Bed

气含率及气泡直径会直接影响流化床内的反应进程及传质效率,为更好地认识局部流动结构,径向气含率及气泡直径分布成为重点研究内容,特别是径向气含率预测模型的建立具有重要意义.已有平均气含率预测模型都有各自的适用范围,对不同实验体系的应用会表现出局限性.针对以上问题,本文结合电导探针测量法与流体仿真技术,以直径为100 mm和高度为1.5 m的气液固流化床为研究对象,系统地研究了在气相表观速度为0.014~0.283 m/s、液相表观速度为0.007~0.028 m/s、液相黏度为1~40 mPa·s、液相表面张力为0.053~0.072 N/m及固相体积分数为0~30%条件下径向气含率及气泡直径分布特征.结合本实验系统及相关研究数据,提出一种适用于气液固系统的平均气含率预测模型.基于此预测模型,进一步考虑气相表观速度及表面张力对径向气含率分布的影响,采用粒子群优化算法建立气液两相体系的径向气含率预测模型.结果表明:两相及三相体系平均气含率预测模型充分考虑了气液相表观速度、液相物性、固相物性及管径对平均气含率的影响,相比现有经典平均气含率预测模型,拓宽了适用范围并提高预测精度,其平均绝对百分比误差(mean absolute percentage error,MAPE)分别为17.14%和18.36%,超出90%数据点的相对误差在±30%之内.径向气含率预测模型对靠近管道中心处的气含率预测更加准确,其MAPE可达到12.26%.以上预测模型的建立对理解管内全局及局部传质问题、流化床反应器设计都具有指导意义.

The gas holdup and bubble diameter directly affect the reaction progress and mass transfer efficiency in the fluidized bed.To better understand the local flow structure,the radial gas holdup and bubble diameter distribution have become the focus of research,with the radial gas holdup prediction model demonstrating great significance.The existing average gas holdup prediction models have a limited scope of application in different experimental systems.Therefore,combined with the conductivity probe measurement and fluid simulation,a gas-liquid-solid fluidized bed with a diameter of 100 mm and a height of 1.5 m was considered the research object in this study.The gas holdup and bubble diameter distribution were systematically investigated under the superficial gas velocity of 0.014—0.283 m/s,superficial liquid velocity of 0.007—0.028 m/s,liquid viscosity of 1—40 mPa·s,liquid surface tension of 0.053—0.072 N/m,and solid phase volume fraction of 0—30%.Combined with the results of this experimental system and literature,a prediction model for the average gas holdup of gas-liquid-solid system was proposed.Based on this prediction model and considering the influence of superficial gas velocity and surface tension on the radial gas holdup distribution,the radial gas holdup prediction model was developed using the particle swarm optimization algorithm.The results indicate that the prediction models for the average gas holdup of two-phase and three-phase systems completely consider the effects of superficial gas velocity,superficial liquid velocity,liquid properties,solid properties,and pipe diameter.The models broaden the scope of application and improve prediction accuracy,with the mean absolute percentage error(MAPE)of 17.14%and 18.36%,respectively.Over 90%of the data points are covered by the new prediction model with a relative error of less than 30%.The radial gas holdup prediction model predicts the gas holdup near the pipeline’s center more accurately,with its MAPE reaching 12.26%.The above prediction models can prove significant in understanding the global and local mass transfer problems in the tube and the design of fluidized bed reactors.