SCOPEThe importance of axial mixing on mass transfer has increasingly been recognized. Fluid mechanical phenomena limit column performance when mass transfer rates are high. Methods for correcting the calculated height or number of transfer units for axial mixing require a knowledge of axial mixing Peclet numbers. Earlier studies on spray columns provided insights into axial mixing but gave no useful correlations for estimating the axial mixing Peclet numbers.New data on dispersion characteristics and flow behavior of dispersed phase droplet swarms have now been reported (Vedaiyan et al., 1972(Vedaiyan et al., , 1974. These studies demonstrated the changing pattern of drop size distribution in the swarms (before flooding). Studies of axial mixing in the continuous phase (Kreager and Geankoplis, 1953; Mixon et a]., 1967; Letan and Kehat, 1968; Henton and Cavers, 1970; Henton et a]., 1973) indicate that axial mixing is primarily caused by drop movement and by carry-over of continuous phase fluid elements in the larger drop wakes. The effects of continuous and dispersed phase Row rates on axial mixing dispersion coefficients E , of the continuous phase are still uncertain. Some analyses assume uniform distribution of drop sizes for all dispersed and continuous phase Row rates, while others assume a constant mean drop size, that is, at constant dispersed phase nozzle velocity. Recent reports of Ve daiyan et al. (1972, 1974) show variations in the drop size distribution and mean drop size related to the velocity at the dispersing nozzles.These results indicate that the continuous phase residence time distribution variance depends, to a great extent, on the same factors which affect the dispersed phase residence time distribution variance and, in addition, on the continuous phase velocity. CONCLUSIONS AND SIGNIFICANCEConclusions from this study are presented under three subheadings as follows. Data analyses folIow methods outlined on Table 1 (after Levenspiel, 1972).1. Conchions from the dispersed phase studies. The dispersed phase RTD variance ,d2 varied significantly with holdup and dispersed phase velocity at the distributor nozzle U N and showed a remarkable similarity to the variation of the drop size distribution variance with nozzle velocity. (The drop size distribution was measured by photographic methods.) It seems reasonable to suggest, therefore, that the factors affecting the residence time distribution of the dispersed phase are the same as those affecting the drop size distribution.The variance of the dispersed phase residence time distribution and axial mixing Peclet numbers were found to be best correlated with dispersed phase holdup rather than with either nozzle velocity or any modified -form of the Reynolds number (see Figures 5 and 6).2. Conclusions from the continuous phase studies. The variance of the continuous phase residence time dis tribution u,2 was found to be a strong function of dispersed phase velocity and holdup. The continuous phase axial dispersion coefficient E, remains nearly const...

Experimental data on the drop size distribution obtained from multiple nozzles in a spray column are presented and analyzed. For nozzles having a diameter above the critical size, the distribution pattern change from near normal monomodal to bimodal and back again to monomodal but skewed toward the larger size fraction. There is no change in the observed maximum drop size in the distribution with flow rate except near the critical velocity region. Drop breakup mechanisms appear to be different from the above pattern for nozzles below a critical size.A model is proposed for prediction of the complete range of drop sizes observed.The performance of a spray-type, liquid-liquid extraction column is dependent on the characteristics of the dispersion. This paper reports results from experiments conducted in the jetting range of nozzle velocities; a photographic analysis immediately above the nozzle was used to obtain the data reported. A model is proposed to characterize the drop size distribution from jetting to disruption velocities, this range being important in the normal operation of spray columns.Relatively few photographic studies of drop size distributions have been reported and the available data are thus meagre. A detailed literature survey is available in the thesis of one of the authors (26) but the following references are particularly noteworthy. Keith and Hixson (11) noted that the uniformity of drops is related to the jet length. They observed that drops produced in the varicose region showed a high degree of uniformity when the jet length is below the critical value. In the sinuous region a decrease in uniformity was observed with an increasingly wide range of drop sizes. They showed the variation of drop uniformity on a normal probability plot of number count and observed that drops formed are most uniform in size near the critical velocity through the nozzle corresponding to the maximum jet length. Hinze (9) studied the maximum stable drop size under shear or turbulent flow, as in a rotary annular column, and suggested that the penetration of lamellae and ligaments of one liquid into another causes disintegration. He suggests that when a long ligament breaks into droplets, secondary smaller droplets are usually formed. Moreover, the ligaments at the moment of breakup will generally not be equally thick and hence, during the distintegration process, drops of different sizes are formed. Christiansen and Hixson (2) stated that the growth of the axially symmetric disturbance causes jet disintegration and that flow rates above the critical velocity produce more nearly random drop distributions. Using high speed photographs of the jets they ob- served that, near the region of instability, a number of nodes exist along the jet profile. Any untimely separation of the nodes upstream produced a drop larger than the ideal one-wave-length drop. The simultaneous disruption of two or more nodes produced small and large drops from the same jet.Weaver et al. (27) photographed drop swarms during the operation of a ...