Study Of The Influence Of Marangoni Effect On Interfacial Area On Distillation Tray

During the design of distillation tray, tray efficiency is usually used to indicate the degree of completion of mass transfer on the tray, so as to obtain the actual plate number according to the theoretical plate number. At present, the theoretical plate number can be already calculated accurately, but how to precisely predict tray efficiency is still very difficult. Because tray efficiency can be affected by many factors, the effect of surface tension has long been neglected. Until the surface tension induced Marangoni effect was found in distillation systems, surface tension gradually gets attention. For distillation systems working in the forth regime, Marangoni effect can alter interfacial area and tray efficiency by inhibiting or promoting bubble coalescence. However, due to the complexity of the distillation system, so far the understanding of the mechanism of Marangoni effect is still a supposition on the macro level, and the its effect on bubble coalescens has not been experimentally verified yet, and without any relevant theoretical models. Thus, making further microscopic mechanism of Marangoni effect is very important for prediction of tray efficiency and design of distillation tray. Therefore, taking the bubble coalescence mechanism in the froth as a starting point, the bubble coalescence behavior with interfacial mass transfer was investigated using experimental and theoretical methods, and theoretical model for liquid film drainage during bubble coalescence was developed depending on whether or not the transferring component has surface activity. Finally, combined with the developed film drainage model and population balance model (PBM), a three-dimentional CFD-PBM model was developed to predict the hydraulics, mass transfer and bubble size distribution on the distillation tray.(1) First of all, an experimental apparatus was built to investigate the interfacial mass transfer effects on bubble coalescence behavior at room temperature. By comparing the bubble coalescence efficiency of several volatile aqueous solutions under different temperatures and concentrations, we, for the first time, verified the supposition of the mechanism of Marangoni effect on bubble coalescence reported in literature on the micro level. Experimental results show that: as the increase of solution temperature, concentration of volatile component and surface active of volatile component, the inhibition effect of Marangoni effect on bubble coalescence is increased, resulting in decreased bubble coalescence efficiency. By using the so called M-index, the intendisy of the inhibition of Marangoni effect of different solutions can be well described.(2) Secondly, focus on the non-surface active components, we, for the first time, added interfacial mass transfer induced Marangoni effect into the film drainage model. The simulating results indicates that, for positive systems, Marangoni effect inhibites film thinning, while for negative systems, Marangoni effect promotes film thinning. For positive system, the drainage process undergo three stages by the inhibition of Marangoni effect, and the Marangoni effect dominates during the second stage. The duration of Marangoni effect depends on the migration rate of the maximum interfacial concentration gradient, which increases with increasing intensity of interfacial mass transfer, increasing diffusion in the film and decreasing approach velocity.(3) Thirdly, in view of the large difference between the surface concentration and the bulk concentration when the transferring component has surface activity, the transport equation of interfacial concentration of the transferring component was added in the governing equations. When the transferring component is surface active, results indicates that:the thinning rate are controlled by both mass transfer induced Marangoni effect and surface activity induced Gibbs-Marangoni effect, resulting in stronger inhibiting effect and lower thinning rate; the drainage process can still be divided into three stages; even though the gas side mass transfer is weak and the mass transfer induced Marangoni effect vanishes, the drainage process is still inhibited by Gibbs-Marangoni effect. Compared with experimental data, the maximal error of the simulating results is below 20%, which verify the reliability of the developed model, but it should emphasize that this model is not applicable for drainage process whose thinning rate is controlled by other factors, such as viscous force.(4) Finally, using the established bubble coalescence film drainage model and basing on the Euler-Euler two-fluid model and PBM model, a three-dimentional CFD-PBM model was developed to predict the hydraulics, mass transfer and bubble size distribution in the froth, through which tray efficiency was obtained. A turbulence model based on the k-E equations for the mixture of the two phases was used to close the N-S equation, and the momentum and mass interchange between phases were obtained by Krishna and Higbie method. By comparing the simulation results with the experimental data in the literature, we found:when the Marangoni effect is not considered, the maximal error of the simulating results of tray efficiency are 33%, and the errors are mainly located in the low concentration region where the Marangoni effect is weak; after the consideration of Marangoni effect, the maximal error decreases to 20%. In addition, accirdubg to the bubble size distribution, bubble break up meanly happens near the tray floor, and bubble coalescence dominates in the bulk froth. When the Marangoni effect is weak, bubble coalescences with each other and becomes large during its rise in the froth, resulting in wide bubble size distribution and a low interfacial area; on the contrary, when the Marangoni effect is strong enough to inhibit bubble coalescence, bubble size keeps almont constant during its rising, resulting in narrow size distribution and a high interfacial area.