节点文献

加压下规整填料塔内流体流动和传质特性的研究及其计算流体力学模拟

Experimental Studies and CFD Simulations of Fluid Flow and Mass Transfer in a Structured Packed Column at Elevated Pressure

【作者】 张鹏

【导师】 余国琮;

【作者基本信息】 天津大学 , 化学工程, 2002, 博士

【摘要】 本文首先在内径为150mm的精馏塔内对Mellapak 250Y型波纹板规整填料在加压下的传质性能进行了测试实验的操作压力为0.3-2.0MPa实验物系为正丁烷/正戊烷实验是在全回流条件下进行的研究表明气相总传质单元高度(HTUOG)可分为两部分一部分为不考虑返混影响的净气相总传质单元高度(HTU*OG)另一部分为由于流体的返混而引起的返混单元高度(HBU)根据填料塔不同轴向位置处液相样中正丁烷的浓度可确定HTUOG和HBU的值并给出了求算HBU的经验关联式结果表明在高压下填料的传质效率随压力的升高而下降气相总传质单元高度从1.017MPa时的0.3810m增大到1.924MPa时的0.5454m 将上述的精馏塔改填为Mellapak 350Y型规整填料来测量气液单相和气液两相流条件下气液相的轴向返混系数实验的操作压力仍为0.3-2.0MPa气液两相在填料塔内逆流流动气相为氮气液相为水它们的流量分别为2.0-5.0m3h-1和0.3-1.2m3h-1(实验条件下)流体的返混特性用示踪响应技术进行测量气相示踪剂为氢气液相示踪剂为饱和的氯化钠水溶液分别用热导法和电导法来检测通过获得的流体停留时间分布曲线采有时间域最小二乘法用一维轴向扩散模型对其拟合可确定不同实验条件下流体的轴向返混系数和有效流速结果表明气相轴向返混系数随有效气速液相喷淋密度和压力的增大而增大但液相喷淋密度和压力的影响要远小于气速的影响液相轴向返混系数随有效液速和气相动能因子的增大而增大压力对其没有影响最后给出了不同条件下计算轴向返混系数的关联式通过引入表征体元的概念采用体积平均的方法建立了规整填料塔内气液单相流和气液两相流的基本输运方程与传统的输运方程相比方程中增加了空隙率相间相互作用力和动力弥散系数三项根据所建的模型用商业流体力学软件PHOENICS3.3对规整填料塔内流体的流动行为进行了模拟分别给出了压力场速度场和示踪剂浓度场的模拟结果根据模拟的示踪剂浓度分布对气液相单相流的轴向返混系数进行了估算所得结果与实验值的平均偏差为22

【Abstract】 The performance of Mellapak 250Y corrugated structured packing in distillation applications at pressures ranging from 0.3 to 2.0MPa had been analysed by using HTU-NTU method. These data were obtained in a 150mm diameter distillation column operated with n-butane/n-pentane system at total reflux. In considering the axial backmixing effects, the height of an overall gas phase transfer unit, HTUOG, was divided into two parts. One part represented the height of an overall gas phase transfer unit, if no backmixing occurs, designated as HTU*OG, and the other part, designated as the height of a backmixing unit (HBU), represented the backmixing effects. The HTUOG was evaluated from the measured concentration profile of n-butane in liquid phase. The HBU obtained experimentally was correlated in terms of the properties of the separating materials and the equivalent diameter of the structured packing. Our result showed that HTUOG varied from 0.3810 to 0.5454m with pressure increasing from 1.017 to 1.924MPa. It indicated the overall efficiency of the structured packing decreased gradually at high pressure, as a result of the vapor backmixing. An experimental study of the extent of axial backmixing in both gas and liquid phases was conducted in a 150mm I.D. column packed with Mellapak 350Y corrugated structured packing. The column was operated at pressures ranging from 0.3 to 2.0MPa with nitrogen and water flowing countercurrently through the packing under trickle-flow conditions. The gas and liquid flowrates were varied from 2.0 to 5.0 m3h-1 (at experimental conditions) and from 0.3 to 1.2 m3h-1 respectively. The amount of axial backmixing was experimentally evaluated by the pulse response techniques using hydrogen in the gas phase and an aqueous solution of NaCl in the liquid phase as inert tracers. The response of the tracer was monitored by means of thermal conductivity in the gas phase and electrical conductance in the liquid phase. The experimentally determined RTD curves were interpreted in terms of the diffusion-type model. The model parameters (backmixing coefficient and interstitial velocity) were determined by the time domain analysis of the response curves. The results indicated that axial backmixing in the gas increased notably with gas flowrate and slightly with operating pressure and liquid flowrate. And the liquid-phase axial backmixing was an increasing function of both gas and liquid flowrates and insensitive to pressure. Various correlations were developed for reproducing the experimental mixing data obtained under one-and two-phase flow conditions. The agreement between experimental and correlated data appeared to be acceptable and within ±20% of difference. According to the definition of the Representative Elementary Unit (REU), the volume averaging technique was applied to derive the governing flow equations under one-and two-phase flow conditions. Additional terms appeared in the averaged governing equations were porosity, interphase forces and hydraulic dispersivity. Depending on these governing equations, a commercial CFD (Computational Fluid Dynamics) code (PHOENICS3.3) was used to predict the fluid dynamics behavior of gas and liquid phases. The simulated profiles of pressure, velocity and concentration of the tracer were presented. Then, the CFD results had been used to fit the diffusion-type model and the obtained axial backmixing coefficient had an average error of 22% with that evaluated by the experimental data.

  • 【网络出版投稿人】 天津大学
  • 【网络出版年期】2006年 11期
节点文献中: