Fluid Dynamics and Diffusion Calculations

for Laminar Liquid Jets

1. E. SCRIVEN and R. 1. PIGFORD
University of Delaware, Nework, Delaware

In connection with a study of the mechanism of gas absorption the problem arose of
predicting absorption rates into laminar liquid jets. A solution to the problem is presented
in this paper, which provides an example of the application of fluid dynamics to the analysis
of mass transfer in a complex flow system.
The water jets considered here issued from circular nozzles of about 1.5-mm diameter,
flowed intact downward through an atmosphere of solute gas at average velocities of
from 75 to 550 cm./sec. over distances of 1 to 15 cm., and were collected in a receiver
slightly larger in diameter than the nozzles. Equations describing the liquid flow near the
jet surface are deduced from measurements of jet diameter and analogy to related flow
situations. When one uses these equations, absorption rates are predicted from unsteady
state diffusion theory with the assumption of interfacial equilibrium. The predicted rates
for carbon dioxide at 25°C are in close agreement with experimental determinations over
the observed range of contact time of the liquid with gas, namely 0.003 to 0.04 sec.
(4, 5, I S , 18) have overlooked these
complications, with the exception of
Laminar liquid jets possess several except that end effects at the points at Cullen and Davidson (2), who discussed
attractive advantages over other types which the film is brought into and out qualitatively the influence of gravity
of apparatus for fundamental studies of of contact with the gas seem unavoid- insofar as it affects the axial velocity
the mechanism of gas absorption. To able, although they may be minimized distribution and who evidently suc-
obtain a valid experimental test of un- t o a certain extent through apparatus ceeded by judicious nozzle design in
steady state diffusion theory in a flow design and experimental technique. Un- greqtly reducing the influence of vis-
system, i t is imperative that the fluid fortunately the fluid dynamics of the oft- cosity. Contact area can be evaluated
dynamics of the system be known ac- employed wetted-wall column is usually readily from measurements of jet di-
curately. The area of contadt between complicated by rippling, which may be ameter. A particular advantage of jets
gas and liquid must be known, and either the cause or an indication of mix- is the wide range of liquid contact times
furthermore the nominal time of ex- ing within the falling liquid film (19). which can be achieved with them, down
posure of liquid to gas should be of the Rippling can be greatly decreased by to a few milli-seconds or even less.
order of 0.001 to 1 sec., since exposure the addition of small amounts of surface-
times of these magnitudes are en- active agents to the liquid feed or al-
countered in most contacting equipment together eliminated by reducing the APPARATUS
of practical interest. If there is departure length of the column sufficiently. How- Detailed descriptions have been given
from equilibrium at the interface, the ever the first p r a c t h raises new uncer- elsewhere of the gas-absorption apparatus
effect of an interfacial resistance on tainties on amount of possible interaction in which the jets of interest here were
absorption rates is likely to be ,greater between solute and surfacfant, while produced (16, 1 7 ) . I n brief, a jet of water
the shorter the period of exposure of the second promotes the end effects to issued from a circular nozzle about 1.5-
liquid to gas, simply because the shorter major importance (If). Recently the mm. I.D. flowed intact downward at an
the contad time the smaller the bulk- idea of carrying a liquid film on a moving average velocity of from 75 to 550 cm./sec.
phase diffusional resistance to mass solid surface has been put to use by over a distance of 1 to 15 cm. and was
transfer. Danckwerts and Kennedy (3) in their collected in a glass capillary receiver about
2 mm. I.D. The receivx and a nozzle
Bubbles of gas rising through liquid novel rotating-drum apparatus, which are shown in Figure 1. Between the nozzle
are manifestly unsuitable, for the flow appears to be nearly free of appreciable and receiver the surface of the jet was
regime in the vicinity of the bubble end effects for most operating conditions. exposed to an atmosphere of solute gas
surface is not well understood (1) and Laminar liquid jets, which have been confined within a glass chamber. The
probably is strongly influended by the utilized in surface tension studies for apparatus was similar to that used by
presence of traces of surface-active con- many years ( 1 4 , are well suited for Matsupma (IY), Manogue ( I $ ) , and
taminants (9). Falling droplets of liquid absorption measurements. By employing Eipper ( 6 ) .
likewise suffer the considerable short- long, fast-moving jets the end effects a t It was possible to collect the entire jet
coming of poorly understood fluid dy- the nozzle and the cbllection device can in the receiver for prolonged periods of
operation without entraining any gas
namics (10). A well-behaved moving be reduced to insignificance, but the bubbles in the liquid, provided the nozzle
film of liquid satisfies the requirements, influences of viscous drag in the nozzle and receiver were accurately aligned, the
and gravitational acceleration on flow resistance to flow downstream of the re-
and absorption must be assessed, as ceiver was carefully adjusted, and strong
L. E. Scriven is with the Shell Development
Company, Emeryville, California. they are herein. Previous investigators vibrations were eliminated. Operating

Vol. 5, No. 3 A.1.Ch.E. Journal Page 397

troubles, when they occurred, seemed to nozzle in a vertical jet the velocity the surface of the jet was negligible, (d)
arise from occasional small fluctuations in profile becomes nearly flat, the principal the minor complication arising from the
the liquid feed rate. force acting is gravity, and the variation small rate of change of diameter with
of velocity with jet length is given to a length was also negligible, and (e) gravi-
FLUID DYNAMICS close approximation by [see the discus- tational acceleration was absent. Except
sion of Equation (9)] in the region close to the nozzle, where
Diameter Measurements gravitational forces were small com-
As the first step in analyzing the fluid u = dU,z +
2gLx (5) pared with viscous shear, the last as-
dynamics of the jet, measurements were sumption is a gross oversimplification,
made with a Gaertner traveling micro- TABLE 1 as is shown below. But with it the prob-
scope of the jet diameter vs. length Dis- Aver- lem of analyzing the velocity distribution
across three diameters 60 deg. apart. Flow, tance, Di- age in the neighborhood of the jet surface
Measurements made a t two flow rates cl? x, ametrr, length, became identical with the problem of
have been plotted in Figure 2. cc./sec. cm. D,em. b E s/Ro 1, cm. calculating the velocity distribution in
From these was inferred the velocity 5.02 0 0,1543 a laminar wake behind an infinitely
distribution. It was assumed that the 5 0.1454 0.103 thin, flat plate oriented parallel to the
velocity within the core of the jet issuing 10 0,1419 0.098 0.088 direction of flow, as depicted in Figure 3,
from the nozzle was uniform and that 15 0.1385 0.098 a problem solved many years ago by
outside the core there existed an annular 2.58 0 0.1546 Goldstein (7, 8). The analogy between
boundary layer in which the velocity 3 0.1406 0.114 the laminar jet and Goldstein's problem
5 0,1362 0.095 0.043 was pointed out by Rideal and Suther-
was reduced, in consequence of the 7 0.1320 0.099
frictional drag exerted by the inside land (14).
surface of the nozzle. I t was supposed For velocities close to the trailing edge
that after the jet left the nozzle, momen- From this expression and the equation of the flat plate Goldstein used three
tum interchange betwcen the core and of continuity it follows that terms of a series solution obtained by
boundary layer would have produced a continuation of the classical Blasius
substantially flat velocity profile beyond solution for a flat plate; for velocities far
some not-too-distant downstream point. downstream he obtained first and second
By assuming that the velocity distri- approximations to an asymptotic solu-
bution within the boundary layer a t the The agreement of observed variations tion:
nozzie could be adequately represented of diameter with the behavior predicted
by a cubic equation, one gets by Equation (6) is quite good at down-
stream distances beyond 2.5 cm. (broken
-U =3 (x) (J
- 1 Y 3, curves in Figure 2).
u, 2 6 In passing i t should be pointed out
that the calculated values of boundary-
O<Y<6 layer thickness do provide confirination
of the laminar character of the flow
_-li in the emerging jet. The Reynolds
- 1, 6 I Y 5 R, (2) number based on boundary-layer thick-
li,
ness U,6/v varies from 100 to 250 for
and writing mass and momentum the jets. I n contrast the analogous
balances between the face of the nozzle critical Reynolds number for flow over
and a downstream point corresponding a flat plate is in the range 1,500 to 7,000
to a flat velocity profile (Figure 3), one (15); furthermore the stability of lami-
obtains the following approximate ex- nar flow is increased in the presence of
pression relating boundary-layer thick- a favorable pressure gradient as obtained
ness to the measured diameters: with the nozzle design employed in the 2 3 X
-
41
+ 0.13; > 0.125
41 -
present work.
6
b E - = 2.094 Figure 4 reproduces Goldstein's complete
R, (3)
Surface Velocity
curve for surface velocity (velocity down-
To predict absorption rates i t is neces- stream in the plane of the flat plate).
sary to know the velocity distribution At distances downstream of the nozzle
in the liquid a t and close to the surface of more than about 2.5 cm. the velocity
The length of flat plate oriented of the jet. For the relatively short con- profiles within the jets were very nearly
parallel to the direction of flow which tact times of interest in connection with flat, and the flow, insofar as it affected
would be required to produce a boundary the present study, the penetration depth jet diameter, could be approximated by
layer of thickness, if one assumes a of the diffusing molecules is very small the free-fall Equation (5). In the down-
cubic velocity distribution, is given by compared with the jet diameter; there- stream portions of the jets, then, gravi-
(16) : fore debailed knowledge of the flow tational effects were far from negligible
within the interior of the jet is not as assumed in (e) above, and in fact
required. they predominated in these regions. Thus
Since the boundary-layer thickness a t another estimate of the liquid velocity
the nozzle, as calculated above, amounted in the neighborhood of the jet surface
Values of b and 1 calculated from the to only one tenth of the nozzle radius, is givcn by
data presented in Figure 2 are given in it seemed reasonable to assume that (a)
Table 1; it is apparent from the near the core velocity remained unchanged
constancy of b that the velocity profiles by momentum exchange with the bound-
within the jets were indeed very nearly ary layer and (b) the annular boundary
flat after 5 cm. or less of travel. This layer could be treated as two dimen- Precisely speaking, Equation (9) as-
conclusion is borne out by a further sional. In addition it was assumed that sumes a strictly one-dimensional flow,
est of the data. Downstream from the (c) the drag of the surrounding gas on although the very slight taper of the

Page 398 A.1.Ch.E. Journal September, 1959

BRASS NOZZLE Also V ( X , 0) = 0, and furthermore L A M I W LIQUID JET
LAMINAR WAFS001WSTREAY
OF THIN FLAT PLATE
U E U, is nearly independent of Y in
+3 0 0 Y * I P P ( I W * LbT"
the vicinity of the surface; hence
0 P M I <CIUUoI)IC*L
T""OAT1

V g - yd U
A (14)
5, ** dX

ABSORPTION

GLLSS RECEIVER Diffusion Equation

The absorption of a slightly soluble,
nonreacting gas is governed by the equa-
UIOm,YLO tion for ordinary diffusion in a binary
system, with constant diffusivity and
mass density:

Fig. I Jet nozzle and receiver. itc

dt
+ (v.V)C = 9V2C (15)
I
jets showed that the flow was in fact I n the jets the penetration depth of the
two dimensional. Because of the small Fig. 3. Velocity Distributions.
absorbed molecules was much smaller
influence of viscosity and the very than the jet diameter; consequently the
nearly uniform jet cross section, how- absorbing liquid can be regarded as
ever, the assumption of strictly one- semi-infinite and its surface as planar. C(0, Y ) = C ( X , m) = C,
(18a, 6)
dimensional flow neglecting d U / d Y is Except in a very small region near the If interfacial equilibrium is assumed and
felt to yield a very good approximation jet surface at the nozzle, diffusion in heat effects at the interface and gas-
for U, far downstream of the nozzle. the axial direction was negligible in phase diffusional resistance are negli-
From the surface velocity one can comparison with convective transport. gible, the third boundary condition is
easily obtain the component of velocity Absorption measurements'with the jets
normal to the surface a t small depths were made during steady state operation. C ( X , 0) = c, (19)
beneath the liquid surface. Continuity Under all these conditions Equation (15)
of an incompressible liquid in two- simplifies to To find a solution of this boundary
dimensional flow requires that value problem, one tries the trans-
dC dC d2C formation C ( X , Y ) = C(q), where
U-+ V-= 9-- (16)
dX dY dY2
71 = Yf(X> (20)
Furthermore d U / d Y was very nearly
But zero a t the jet surface (the adjoining Equation (17) then becomes
gas exerted negligible drag upon the
liquid surface) and beneath to a large
depth as compared with the penetration
depth (16);hence U can be replaced by If f ( X ) satisfies
dU the surface velocity and V can be ex-
= Z d X - -dY (12) pressed in terms of U, by means of
dX dX U,-
- df - __1 d__
U, -
- -29 (22)
Equation (14): f" d X f2 dX
then Equation (21) simplifies to

--;txl
- d YUdY Two boundary conditions appropriate to
the problem are
which, integrated once, gives

A second integration, with Equation

(18), yields

C = C, + A 1'e-" m
dz (25)

The remaining constant of integration

is evaluated by means of Equation (19);
thus

25 50 75 I00 125 150 I75

= erfc (q)
LENGTH, MY
This is the desired solution provided
Fig. 2. Jet diameter for 1.535-mm. I. D. paraffin-coated nozzle. that j ( X ) satisfies Equation (22), a

VOl. 5, No. 3 A.1.Ch.E. Journal Page 399

5
"Q

o og 02 03 0. 0s 06 or oa 0s 10
-X
-X
4L
4L

Fig. 5. Effect of nozzle drag on local absorption rate.

Fig. 4. Surface velocity after Goldstein (7, 8 ) .

Bernoulli equation, the solution of which Now the local absorption rate must
is approach that into an ideal jet as viscous
U,(X) = 4(C, - C,)dBUqh drag in the nozzle and gravitational
(27) acceleration approach zero; therefore
JL 1 U , ( X ) dX
-

f(x)= 453
Approximate Solution for Absorption Rate P(0, 0) is unity. After evaluation a t
An exact prediction of absorption rate Xz = X F = 0, the various partial de-
Hence for any given surface velocity requires accurate knowledge of the ve- rivatives may be replaced by constants,
the reduced coordinate q is given by locity distribution near the surfahe of a say a,, PI, yll, etc., giving
jet affected simultaneously by viscous
n =
I.'I 1,( X )
(28)
drag in the nozzle and by gravitational N
- = -1 + (1 + alxl

+ 53 / U,(Xj dX
acceleration. The combined effect is not N*
d
Z3 accurately known; however the separate
effects can be assessed from the esti-
mates of surface velocity provided by
+ a23ZL2 + . . .) (37)
The constant B is evaluated by means
of Equation ( H a ) , which requires that
q ( X , Y ) -+ 03 in the limit as X -+ 0,
Equations (7) to (9) and can then be
combined to give a fairly accurate pre-
+ (1 + PlDIF + p2XF' f . ")
or else through some other boundary
condition in X.
diction of absorption rate. The method
of combination can be explained as
+ Y11XZXF + . * '

Loch1 absorption rate, rather than follows. (higher-order interaction terms)

the concentration distribution given by On the basis of Equations ( 7 ) , (8),
Equation (26), is required for absorption (9), and (31) the functional dependence Here the first series may be identified
predictions: of local absorption rate can be written as with jet operation in a vanishingly small
gravitational field, that is with FI =
F(&, 0), and the second series may be
N = --%(%) Y=O
identified with operation in the ab-
(29) sence of nozzle drag, that is with F2 =
2 Purely dimensional considerations lead F(0, X F ) .It then follows (provided cer-
= (C, - C<l) %f(X) immediately to a more fruitful, func- tain mathematical conditions are satis-
4
~

tional relationship in terms of dimension- fied) that for sufficiently small values
For an idealized jet throughout which less variables : of the nozzle drag and Froude numbers
velocity is constant (equal to Uo = the local absorption rate is given to a
4 q / r Do*)it follows from Equations (27) good approximation by
and (29) that
N

If the first two dimensionless groups are

I t is convenient to refer absorption rate regarded as fixed parameters in the where first- and higher- order interaction
to this rate for an ideal jet; the ratio is problem, expansion of Equation (35) in terms have been neglected.
a Maclaurin series yields The functions PI and F2 are obtained
N from the estimates of surface velocity
N* (31) -
N*
- F(0'0) + x1 and from Equation (31). From Equa-
tion (7)

F, = .$*
Finally the total absorption rate of
solute gas by a jet of length h is ob-
tained by integration over jet length:
+ 3112
2
(5)0.0
(39)

6= K U ( X ) N ( X >dX (32)
0

For an ideal jet the diametcr as well as

velocity is constant; hence + % (g.),,,
Page 400 A.1.Ch.E. Journal September, 1959
 TABLE
2.

D O = 0.1535 cm., Uo = 271 cm./sec., 41= 0.352 cm., U, = 292 cm./sec. ACTUAL JET PROFILE

0.01 35.2 0.720 0.00013 1.000 0.720 1.000 0.545 0.01

0.1 3.52 0.937 0.00133 1.001 0.938 0.982 0.786 0 . 1
1 0.352 1.026 0.0133 1.010 1.036 0.968 0.916 1
10 0.0352 1.038 0.133 1.091 1.129 0.924 0.988 10
DO = 0.1535 cm., Uo = 145 cm./sec., 41 = 0.172 em., Urn = 150 cm./sec.
0.005 3 4 . 4 0.710 0.00047 1.000 0.710 1.000 0.545 0.005
0.05 3.44 0.923 0.00467 1.001 0.924 0.987 0.762 0.05
0.5 0.344 1.009 0.0467 1.017 1.026 0.966 0.911 0.5
5 0.0344 1.019 0.467 1.149 1.168 0.889 0.993 5

where the constant B of Equation (31) rate is indicated in Table 2, in which

0 I 2 3 4
has been taken as zero in accordance
with Equation (1%). From Equation (8)
representative values are listed for jets
,f-. h''; CM?

a4 a, Fig. 6. Carbon dioxide absorptioninto water.

DISCUSSION

The experimental and predicted ab-

sorption rates agree closely, but this fact
alone verifies neither the approximate
2 = -
41
A
+
0.13; A > 0.125
41 - produced with a 1.535-mm. I.D. brass solution of the diffusion equation from
which the predictions arise nor the
nozzle at flow rates of 5 and 2.6 cc./sec.
where B has been set equal to 0.0956 Equation (32) with Equation (38) was particular set of boundary conditions
(161 DU,) to give a smooth transition integrated numerically to arrive at the used to obtain that solution. Before
from Equation (39) in the region predicted total absorption rates plotted any conclusion is drawn, it is necessary
0.0486 < X/41 < 0.125 (Figure 5). as full curves in Figure 6. The predicted to consider the extent to which the
From Equation (9) approximate solution may deviate from
the exact solution. I n over-all absorption
rate the former leads to predictions
differing from those based on the as-
sumption of an ideal jet by no more
than 10%. Such deviation may be re-
garded as a first perturbation to the
ideal-jet absorption rate, caused by the
I disturbing influences of viscous drag in
24 the nozzle and gravitational acceleration.
The calculation of the parameters Now the principal part of the unknown
I and U o has already been described. absorption rates for an ideal jet, accord- second perturbation is proportional to
The core velocity within the emerging ing to Equation (33), are plotted for the product of the already small dis-
jet was evaluated from the average ve- comparison as broken lines. turbing influences [Equations (37) and
locity, the assumed form of the velocity Experimentally determined absorption (38)] and may reasonably be expected
distribution in the boundary layer, and rates for nominal contact times ranging to be about 1% of the ideal-jet pre-
the inferred boundary-layer thickness at from 0.003 to 0.04 sec. are also plotted diction; the higher-order perturbations
the nozzle by means of in Figure 6. Details of the experimental should be proportionately smaller. It
procedure have been reported elsewhere is therefore quite unlikely that the prc-
(16, 17). I n brief, feed and effluent dictions based on the approximate solu-
water samples were drawn with the tion are more than 2 or 3y0 in error, even
apparatus operating a t steady state. Due under the conditions of the lowest ab-
precaution was taken to prevent transfer sorption rates measured, for which the
of carbon dioxide to or from the samples. deviation from ideal-jet behavior was
Comparison with Experimental Results greatest.
Analysis for dissolved carbon dioxide
By the use of the foregoing equations depended upon the precipitation in basic Thus the data substantiate the physi-
absorption rates at 25°C. of carbon solution of carbonate ion as barium cal model underlying the theoretical
dioxide into several laminar water jets carbonate, followed by back titration of predictions; in particular, they confirm
were predicted. The diffusivity of carbon excess hydroxyl ion with standard acid the validity of the asumption of thermo-
dioxide in water a t 25°C. was taken as to the carbonate-bicarbonate end point. dynamic equilibrium between gaseous
1.97 X 10-5 sq. cm./sec., and the solu- Both the precision and accuracy of the and liquid phases everywhere on the jet
bility a t the same temperature and analysis were found to be about 0.4 surface [Equation (19)]. This matter has
760 mm. Hg partial pressure was taken p.p.m.; the range of concentrations ana- been discussed in detail by the authors
as 3.39 X 10-5 g.-mole/cc. The initial lyzed in the effluent water was 10 to in a n earlier paper (1'7).
concentration, was assumed to be zero, 44 p.p.m. That the reproducibility of
the water having be& carefully stripped replicate runs was, with few exceptions, ACKNOWLEDGMENT
of dissolved gas before use (16). excellent can be seen from the data. Thanks are due M. M. Wcndel and
The order of magnitude of the adjust- The agreement of experimental and pre- Stanley Corrsin for helpful discussions of
ments to the reference local-absorption dicted absorption rates is very good. portions of this work and W. J. Baur for

Vol. 5 , No. 3 A.1.Ch.E. Journal Page 401

assistance with some of the measurements. is small compared with the width of the R = jet radius, cm.
The investigation was made possible in flow and only the first term within the u = axial component of velocity,
part by National Science Foundation and braces need be retained. Then the local cm./sec.
Shell FelIowship Committee fellowships re- absorption rate is given by V = radial component of velocity,
ceived by L. E. Scriven. cm./sec.
N = ( C , - CO) 2 = +
(X/41) 0.13, dimensionless vari-
APPENDIX
9uo able, Equation (8)
The nature of the problem of absorption X = axial coordinate, cm.
into an accelerating liquid flow is illustrated 'JTRo(l - 'IzIn ['/(' - 1' 1 y = radial coordinate, cm,
by a simple example. For a two-dimensional where now x = (Ro - R)/Ro.
flow directed toward a sink a t the origin This problem can also be solved by the Greek Lettee
and bounded by free streamlines at the method presented herein. Instead of the & = boundary-layer thicklless, om.
acute angles f eo in polar coordinates the polar coordinates just used, rectangular TJ = dimensionless variable, Equation
potential and stream functions are, re- coordinates with the origin in the free
spectively, streamline a t RO are employed. With this (20)
transformation the surface velocity is v = kinematic Viscosity, sq. cm./sec.
= RoUoIn R E = (X/41)1/3, dimensionless variable,
Ro uo Equation (7)
k = RoUoO TJ-8 = ___ = total absorption rate, g.-moles/sec.
Ro - X C$

where Uo is the radial velocity a t distance SubscriPfs

Ro from the origin. Such a flow corresponds and the velocity normal to the free stream-
to a two-dimensional liquid jet in a certain line is e = eqdibrium
body force field. If one considers absorption 0 = nozzle exit
along the free streamlines between Ro and V = - RoU,Y s = surface
R, 0 << R < Ro, the appropriate diffusion
problem is
+
(R, - x)2 y" 00 = core
* = ideal jet
For shallow penetration depths Y is small
in the region of interest, and V is given
to a good approximation by LITERATURE CITED
1. Birkhoff, Garrett, "Hydrodynamics,"
Princeton Univ. Press, Princeton, N. J.
(1950).
2. Cullen, E. J., and J. F. Davidson,
which can also be obtained with the use of Trans. Faraday Soc., 53, 113 (1957).
ac
- ( R ,0) = 0
Equation (14). From Equations (27) and 3. Danckwerts, P. V., and A. M..Kennedy,
(29) for local absorption rate, given the Trans. Znst. Chem. Engrs. (London),
dR surface velocity, the result is again found 32, 549 (1954).
where the radial velocity is related to the to be 4. Dirken, M. N. J., and H. W. Mook,
potential function by Biochem. Z . , 219, 452 (1930).
N = (C, - C,) 5. Edwards, G., R. Robertson, F. Rum-
ford, and J. Thomson, Trans. Znst.
Chem. Engrs. (London),32, S6 (1954).
With the usual transformation for diffusion 6 . Eipper, J. E., thesis, Univ. Delaware,
in potential flow, in which the potential Newark (1955).
There is of course no flow across the 7. Goldstein, Sydney, Proc. Cambridge
and stream functions are taken as the streamlines (lines of constant @), but for Phil. Soe., 26, l(1930).
independent variables, and with diffusion Y > 0 there is a velocity component in 8. -, Proc. Roy. SOC.(London),A142,
along the stream tubes neglected, there the direction normal to the interface which
follows 545 (1933).
must be accounted for when the diffusion Y. Hammerton, D., and F. H. Garner,
equation is written in rectangular coordi- Tram. Znst. Chem. Enars.
" (London),
.
nates with the origin in the interface. 32, 518 (1954).
10. Hughes, R. R., and E. R. Gilliland,
NOTATION
Chem. Ena. Prow. Svmposium Ser.1 .

No. 16, 51," 101 (1955).

a = dimensionless constants, Equa- 11. Lynn, Scott, J. R. Straatemeier, and
tions (7) and (8) H. Kramers, Chem. Eng. Sci., 4, 49,
b = 6 / R o , dimensionless boundary- 58 (1955).
layer thickness 12. Manogue, W. H., Ph.D. dissertation,
Univ. Delaware, Newark (1957).
which is virtually identical to the problem B = constant in Equation (27), 13. Matsuyama, Takuzo, Mem. Fac. Eng.
of diffusion into an infinite sheet, the cm.4/sec.2 Kyoto Univ., 15, No. 11, 142 (1953)
solution of which is well known. For small C = concentration, g.-moles/cc. (in English).
values of +a - @ (short exposure times) Ca = initial concentration 14. Rideal, E. K., and K. L. Sutherland,
the local rate of absorption is expressed D = jet diameter, cm. Tmns. Faraday SOC.,48, 1109 (1952).
conveniently as 33 = diffusion coefficient, sq. cm./sec. 15. Schlichting, Hermann, "Boundary
f = function defined b y Equation (27) Layer Theory," McGraw-Hill, New
N = (C, - C'J York (1955).
gL = local acceleration of gravity,
DUO cm./sec.Z 16. Scriven, L. E., Ph.D. dissertation, Univ.
Delaware, Newark (1956).
- J & o ( R w o Y In (Bo/E) h = jet length, cm. 17. ___ , and R. L. Pigford, A.1.Ch.E.
1 = equivalent length of flat plate, cm. Journal, 4, 439 (1958).
N = local absorption rate, g.-moles/ 18. Vielstich, Wolf, Chem.-Zng.-Tech., 28,
(sq. cm.)(sec.) 543 (1956).
X p = gLX/Uo2,Froude number 19. Vivian, J. E., and D. W. Peaceman,
Tip = UoX/33, Peclet number A.Z.Ch.E. Journal, 2, 437 (1956).
Xu = U,/Uo, velocity ratio
If the distance RO - R is sufficiently short, 3& = 4Z/X, nozzle-drag number Manuscript receiaed Julu 22. 1968; revision re-
cehed J a n u a r y 20, 1950; paper accepted January 22,
the penetration depth of absorbed material q = volumetric flow rate, cc./sec. 1959.

Page 402 A.1.Ch.E. Journal September, 1959