Вы находитесь на странице: 1из 3

NOTES The routine now is to measure Usb (by measuring Jg, J~, and %), estimate UT from Eq.

[2] (with m = 2), and iteratively solve for db in Eqs. [3] and [4]. Other expressions for Ux, provided db ~< 1.5 mm, could be substituted (Dobby et al. (1)). One simplification that becomes apparent on use is that there is now only one definition for Re instead of, as before, one for the determination of m and another for the determination of U~b. Using this new routine the data of Yianatos et al. (6) were reexamined; an extract of the results, selected to cover the full range in db, is given in Table I. Essentially no difference with the previous result is found. CONCLUSION The simplified approach, based on a drift flux analysis with m = 2 and an expression for UT applicable to db ~< 1.5 mm, gives adequate estimation of db in the tested range of 0.5 to 1.5 mm, APPENDIX: NOMENCLATURE db g Jg Jl m Re Usb Ux bubble diameter, general term, cm acceleration due to gravity, c m / s 2 superficial gas velocity, c m / s superficial liquid downward velocity, c m / s parameter, Eq. [ 1] bubble Reynolds number, Eq. [4] slip velocity between bubbles and liquid, e m / s bubble terminal velocity, c m / s

299

Greek Symbols
eg fractional gas holdup ol liquid density, g / c m 3 //1 liquid viscosity, g / c m . s REFERENCES 1. Dobby, G. S., Yianatos, J. B., and Finch, J. A., Canad. Metall Q. 27(2), 85 (1988). 2. Masliyah, J. H., Chem. Eng. Sci. 34, 1166 (1979). 3. Richardson, J. F., and Zaki, W. N., Trans. Inst. Chem. Eng. 32, 35 (1954). 4. Schiller L., and Naumann, A., Z. Ver. Dtsch. Ing. 77, 318 (1933). 5. Shah, Y. T., Kelkar, B. G., and Godbole, S. P., AIChE J. 28(3), 353 (1982). 6. Yianatos, J. B., Finch, J. A., Dobby, J. S., and Xu, M., J. Colloid Interface Sci. 26 ( 1 ), 37 ( 1988 ). MANQIU X u J. A. FINCH

Department of Mining & Metallurgical Engineering McGill University Montreal, Quebec Canada H3A 2A 7 Received March 5, 1990; accepted April 13, 1990

Journalof ColloidandlnterfaceScience,Vol. 140,No. 1, November 1990

The So-Called "Ohnesorge Equation"


In a paper published in this journal, Richardson ( 1 ) discussed emulsification by jet injection. The premise of the paper was that the jet velocity had to exceed a critical value Vo in order for the jet to break up, and thus for emulsification to occur. Richardson presented the following equation for the calculation of this critical velocity
rh/(pi'YiD) 1/2 = 2000(rh/VoPlD) 4/3,

[1]

(especially for D = 0.275 ), pointing to, in all probability, a computational error. At this point, it became clear that to make any sense of these results, it would be necessary to refer to Ohnesorge's original paper. Although written in the turgid academic German of its period, it is perfectly clear that this paper does not contain Eq. [1] in any form. Rather, Ohnesorge introduces a dimensionless group
Z = o/(3,ioD) 1/2,

where D is the nozzle diameter, pl and ~/1 are the density and viscosity, respectively, of the internal phase (i.e., of the liquid being forced through the nozzle), and 3'i is the interfacial tension. Richardson attributes this equation to Ohnesorge (2), and it has been referred to in the literature as the "Ohnesorge Equation" (3). The critical velocity V0for jet breakup is the value of Vo which satisfies Eq. [1], and Richardson ( 1 ) gives a number of such calculated values. Recently, having become again interested in the phenomenon of jet breakup, I reinvestigated this work. To my surprise, I could not duplicate Richardson's calculation of the critical velocities. In fact, some of the values appear to be quite different (cf. Table I). Although the values for D = 0.1 cm and D = 0.15 cm are fairly close, Richardson's values for D = 0.275 cm differ rather strongly from my calculations. Furthermore, Richardson's values are evidently rounded, but the extent of the rounding is unknown (presumably Richardson used logarithms to do his calculation). However, from Eq. [1] the dependence of the jet velocity on the diameter is as D5/8; thus, for a given set of liquids, the velocities should vary as 1:0.776:0.531 for the diameters under consideration here. While our results meet this test, Richardson's fail rather substantially

[21

where the quantities are defined as in Eq. [1]. Z is known as the Ohnesorge Number, and Eq. [ 1] may now be written Z X Re 4/3 = 2000. [3]

Although Ohnesorge implies a relationship between Z and Re, there is nothing in Ref. (2) to suggest anything of the form of Eq. [ 3 ]. Ohnesorge does present a log-log plot of Z vs Re (which, unfortunately, cannot be reproduced here), showing experimental points corresponding to various liquid pairs. The plot is divided into three zones corresponding, respectively, to I: droplet breakup according axisymmetric surface oscillations (Rayleigh oscillation); II: breakup by screw-symmetric oscillation; and III: atomization. The three zones are delimited on this plot by two parallel lines with a slope o f - 4 / 3 . Hence, the equation of these lines is Z = f(Re-4/3). [4]

The factor 2000 (in Eq. [l]) is nowhere to be found in Ohnesorge, and its origin is a mystery (except, of course, for the fact that Re = 2000 is usually regarded as a critical

TABLE I Critical Velocities from the "Ohnesorge Equation" D-0.1 Liquid Aniline-water Benzene-water Paraffin-water From (1) 70 100 150 This work 66.0 90.5 150.2 From (1) 50 70 110 D=0.15 This work
51.2 70.2 116.6

D - 0.275 From (1)


19 25 90

This work
35.1 48.1 79.8

300 0021-9797/90 $3.00 Copyright 1990by AcademicPress,Inc. AI1rightsof reproductionin any formreserved.
Journal of Colloid and Interface Science, Vol. 140,No. 1, November1990

NOTES value), and Ohnesorge, while emphasizing the effect of the jet velocity, never specifically mentions a "critical velocity." It should he noted that the values of Z and Re corresponding to the data of Table I put all the systems into the first Ohnesorge category I (Rayleigh instability), which, I cannot believe, is what Richardson intended to demonstrate. Both Richardson (1) and Ohnesorge (2) present photographic evidence that purport to show the effect of the critical velocity. These photographs, although striking, do not make a quantitative point. Perhaps it would be well, in the future, to call Eq. [1] "Richardson's Equation." New experimental data might be of value. REFERENCES

301

1. Richardson, E. G., J. ColloidSci. 5, 404 (1950). 2. Ohnesorge, W. v., Z. Angew. Math. Mech. 16, 355 (1936). 3. Becher, P., "Emulsions: Theory and Practice," 2nd ed., p. 279. Krieger, Melbourne, FL, 1977 (reprint of 1966 ed.). PAUL BECHER
Paul Becher Associates Ltd. Wilmington, Delaware Received March 5, 1990; accepted April 25, 1990

Journal ofColloidandInterfaceScience, Vol. 140,No. 1, November1990

Вам также может понравиться