Bubble Formation at an Orifice in an Inviscid Liquid .pdf
File information
Original filename: Bubble Formation at an Orifice in an Inviscid Liquid.pdf
This PDF 1.6 document has been generated by Adobe Acrobat 11.0.20 / Adobe Acrobat Pro 11.0.23 Paper Capture Plugin, and has been sent on pdfarchive.com on 06/03/2020 at 20:19, from IP address 152.89.x.x.
The current document download page has been viewed 360 times.
File size: 1.2 MB (8 pages).
Privacy: public file
Download original PDF file
Bubble Formation at an Orifice in an Inviscid Liquid.pdf (PDF, 1.2 MB)
Share on social networks
Link to this file download page
Document preview
BUBBLE FORMATION AT AN ORIFICE IN AN
INVISCID LIQUID
By J. F. DAVIDSON, M.A., Ph.D., A.M.I.Mech.E.* and B. 0. G. SCHULER, Ph.D.*
SUMMARY
The periodic formation of bubbles due to the flow of gas into an inviscid liquid has been studied both experimentally
and theoretically. Two systems were studied:
(i) The simplest case is when gas flows through an orifice into the liquid at a constant volumetric rate, so that
the movement and growth of the bubble do not influence the gas flow rate. In this case a simple theory, based on
first principles, gives a rough estimate of the relation between bubble volume and flow rate.
(ii! When gas passes into the f?rming bubble fr.om an infinite vessel at constant pressure, the theory has to be
modified to allo~ for the change m gas !low rate mto the bubble as it forms. Nevertheless, the relation between
gas flow rate, G, and bubble_volume V is similar to the relation for case (i). However, when the pressure in
the vessel is reduced, G and_V are ~uddenly reduced to zero at a certain value of the vessel pressure, so that
there are cri~cal val~es of G and V. Th~~ critical value~ of pressure, flow, and bubble volume are in rough
agreement mth expenmental valnes, and It is hoped that this theory will throw some light on the phenomenon of
"dumping" on sieve plates.
Introduction
This paper describes an experimental and theoretical investigation of the formation of gas bubbles at an orifice in an
inviscid liquid. The main purpose of the paper is to describe
a theory of bubble formation by means of which the size and
frequency of bubbles can be calculated. This attempt has
been only partly successful in that the size and frequency of
the bubbles do not always agree with experiment, but the
theory is nevertheless valuable in assessing the approximate
effect of the various parameters. In particular, the theory is
valuable in predicting that there is a critical gas fl.ow rate for
continuous bubble formation in the socalled "constant
pressure" experiments. In these experiments, air was passed
through a single hole of radius r0 in a horizontal plate into a
finite depth of liquid above the plate. It is clear that for
bubbles to form, the pressure P 1 in the vessel below the plate
must exceed a minimum value equal to the sum of the static
head of liquid h above the plate plus a surface tension term
2a/r 0 • The theory showsand experiments confirmthat if
the vessel below the orifice plate is large, then as soon as P 1 is
greater than pgh + 2a/r0 , bubbles of a finite size will form
continuously, so that the gas fl.ow rate suddenly increases from
zero to a finite amount. Here p is the liquid density and g the
acceleration of gravity. This finite gas flow is of the same order
of magnitude as the critical gas flow per hole on a sieve plate,
the socalled "dumping rate". 1 • 2
The paper also deals with studies of bubble formation with
a constant gas fl.ow rate, the constant flow being obtained by
feeding the air into the liquid through a restriction such as
a capillary or sinter, so that there is a large pressure drop
between the gas reservoir and the feed to the bubble. This
arrangement is of less practical importance, but is simpler to
treat theoretically and is therefore useful in assessing the merits
of the theory.
The present work is a continuation of work described in a
previous paper on bubble formation in a viscous liquid. 3
Numerous investigators4  16 have measured the volume of
bubbles formed when air or other gas is blown into an inviscid
liquid. These experiments fall into three classes:
* University of Cambridge, Department of Chemical Engineering,
Pembroke St., Cambridge.
TRANS. INSTN CHEM. ENGRS, Vol. 38, 1960
(i) With a very low gas rate4 • 6 • 6 • 7 • 8 the mechanism of
bubble formation is similar to the mechanism of drop
formation in the "drop weight". method of determining
surface tension. These gas flow rates are much less than
those considered in the present paper.
(ii) When the bubbles were formed at the end of a tube
a few millimetres in diameter6 • 9 • 10 • 11 or at an orifice
connected to a small gas buffer vessel, the conditions
approached those of constant gas fl.ow during bubble
formation.
(iii) Where the gas buffer vessel was larger, 12 16 the
conditions were more nearly those of constant pressure,
though with an intermediate sized vessel, the volume of the
vessel is an important factor determining bubble size.
No previous attempts appear to have been made to predict
bubble volumes and frequencies from theory, though Hayes,
Hardy, and Holland16 have written down equations describing
the vertical motion of the bubble during formation.
Experimental
Apparatus for bubble formation
The apparatus has been described in detail in a previous
paper. 3 The liquid in which bubbles were formed was contained in a vertical tube of 14·7 cm internal diameter with
liquid depths up to 15 cm. This tube was mounted with its
axis vertical, on top of a 45litre drum which acted as a
buffer, gas being fed steadily into the drum and passing out
as bubbles through an orifice at the base of the cylinder
containing the liquid. Orifices of three kinds were used, and
are illustrated in Fig. 2 of the previous paper: 3
(a) For the constant pressure experiments, the orifice
was essentially a plain hole in a horizontal plate with a
diameter between 0·29 and 0·46 cm.
(b) For the constant flow experiments with large gas
fl.ow rates, the orifice consisted of two plates. The top plate
was in contact with the liquid and was drilled with a plain
hole with a diameter between 0·30 cm and 0·50 cm. The
bubbles formed immediately above this hole, and to ensure
constant flow a sintered plate was brazed on below the
hole.
336
DAVIDSON AND SCmJIBR. BUBBLE FORMATION AT AN. ORIFICE IN AN INVISCID LIQUID
(c) For the constant flow experiments with smaller fl.ow
rates, the gas was passed into the liquid through a long
capillary. The end of the capillary where the bubbles formed
was let into a horizontal plate so that the bubbles were
formed under conditions similar to those with orifices (a)
and (b) above.
The arrangements for measuring gas flow and drum pressure were described previously. 3
In most of the experiments the bubble frequency was
measured by a stroboscope. In some of the constant pressure
experiments, the bubbleformation was not regular enough
for the use of a stroboscope, and it was necessary to take
cine photographs to measure the volumes of the individual
bubbles. For this purpose a Pathe Webo cine camera was
available, with up to 80 frame/s, and a minimum exposure
time of about 1/600 s.
Constant Flow
Theory
The assumptions used in deriving the theory are similar to
those described in the previous paper: 3 the bubble is assumed
to be spherical at all times during formation, and the upward
motion is determined by a balance between the buoyancy
force and the upward mass acceleration of the fluid surrounding the bubble. Thus the equation of upward motion is:
Vg
d(1116 V dtds)
= dt
s
= 16g
11
[!: + V
0t
4
2G
_
Vo 2 l (Gt
2G 2 n
+
Vo)]
Vo
Experimental results
The results are classified according to the flow rate; when
this is small, equation (2) is appropriate, and equations (3)
·and (4) are used for larger flow rates.
Fig. 1 shows the results for small flow rates compared with
equation (2). It will be seen that above a flow rate of 1·5 ml/s
the results are in excellent agreement with theory. At the
lower flows the discrepancy between theory and experiment is
greatest for the larger orifice with air bubbling into water.
Approximate calculations suggested that the discrepancy is
00 8
0
(1)
G6/s
(2)
As each bubble breaks away from the orifice it leaves behind
a volume of gas V0 which forms a nucleus for the next bubble.
In the derivation of equation (2) it was assumed that Vo was
much less than V1 , but with larger orifices V0 is not negligible
and it is necessary to put
4
V = j rcr 3 = Gt + V 0
(3)
. (4)
The bubble is assumed to detach whens= r + ro, and the
lifetime of the bubble is obtained by plotting equation (4),
and plotting r + r 0 as a function of time from equation (3) on
the same axes. Here r 0 is the radius of the orifice supplying
the gas and it is assumed that V 0 = 4nr 03/3.
0·07
.
V being the volume of the bubble at time t after it started
growing, ands the vertical distance of the centre of the bubble
above the point where the gas enters the liquid.
The assumptions and terms neglected in deriving equation
(1) are:
(i) The fl.ow around the bubble is assumed to be irrotational and unseparated, and the effective inertia of the
surrounding fluid has been taken as 1lp V/16. 3 Because of
this assumption, the drag coefficient is zero. Hayes, Hardy,
and Holland16 assigned a drag coefficient to a forming
bubble, thereby assuming that there is a fully established
wake behind each bubble as it forms. However, when a
solid sphere or cylinder is accelerated from rest in a fluid,
the initial motion is practically irrotational, and the wake
is not fully established until the body has moved an appreciable distance. 17 • 18 It seems likely that the flow round
the accelerating bubble will be similar, and consequently
the drag coefficient is likely to be negligible until after the
bubble has detached from the orifice.
(ii) Neglect of the upward momentum of the gas leaving
the orifice is justified by calculations previously described. 3
With a constant gas flow G, and an initial bubble volume
ofzero, V = Gt; by substituting this relation in equation (1),
the latter can be integrated. When t = 0, ds/dt = 0, and
assuming that the bubble growth is terminated when its
radius r equals s the distance travelled, gives the final bubble
volume 3
V1 = l ·378 ga/s
when integrating equation (1). This gives, with the initial
conditions s = ds/dt = 0,
0 ••
0 05
l
"':>
0·04
"
.J
0
>
"'
"'"':>
"'
.J
8
003
00
rP
p
0·02
P<t.Ethu 0
0·00J
tar
o·s
I 0
v
ro
(c)
0 ·81
0·99
27·1
·0334
l·O
1·0
72·7
72·7
·033"6
·0143
Wotcr
o
1 ·0
Wol<r
)(
I ·O
t·378
c 1 2 1.0 ·6
,. 5
a
(9/ml)
2·0
(dyn /cm) (cm)
2·5
3·
I)
fig. I .Experimental and theoretical bubble volumes for small constant
flow rotes
due to the influence of surface tension acting round the rim
of the orifice, as it does in "drop weight" experiments, an_d
of course as the flow rate tends to zero the bubble volume is
determined by the balance between surface tension and
buoyancy forces, as in the "drop weight" method.
Fig. 2 shows the results for higher flow rates, compared
with theoretical curves calculated from equations (3) and (4)
using various values of r 0 • These theoretical curves give bubble
volumes that are somewhat higher than those given by equation (2). With these higher fl.ow rates. the double and quadruple bubble formation observed by Helsby and Tuson19 was
seen to occur. This made thr. stroboscopic method impracticable, and to measure the bubble volume it was necessary to
TRANS. INSTN CHEM. ENGRS, Vol. 38, 1960
= 41(12)
n
1/5
1378
i1
3/5
DAVIDSON AND SCHULER. BUBBLE FORMATION AT AN ORIFICE IN AN INVISCID LIQUID
take cine pictures of the forming bubbles. The volume of
each individual bubble could be calculated from the pictures,
and for comparison with theory Fig. 2 shows the volume of
the leading bubble in the series of two or four, before coalescence.
..
,
0
0
r 0 0·25cm
::
r0
..
0·2ocm
r0
'"'
0·15 cm
x •
337
range 3·015 ml/s but there is good reason to suppose that
theory and experiment would agree in this range. At fl.ow
rates above 20 ml/s the divergence between theory and experiment is believed to be partly due to the upward current
induced by the bubbles in the liquid surrounding the orifice
and partly due to the deformation of the forming bubbles
from the spherical shape. Because of the upward current,
each forming bubble is dragged upwards and therefore has
a smaller volume than the theoretical bubble which is imagined
to form in a stagnant liquid. Also, the flattening of the base
of the forming bubbles may cause them to detach earlier than
the idealized spherical bubbles.
I 0
Consianl Pressure
~
Theory
w
""....
0
>
w
~
p
ii:"' o· s
Weter
Wotcr
•x
(g/nil)
(c)
0
'o
(dyl'l/c.m) (cm)
rs
I 0
I 0
72·7
0
I 0
l·O
72·7
020
0
0
I 0
10
72·7
0 2'
Aq. Gyurol
I 17
11·1
68·2
0·20
Pel Ethcr
/:,
0· e.1
0·99
27· I
0' 20
Wolcr
Ttllory
,,
20
G ·GAS
The upward motion of the forming bubble is determined by
equation (1) in the same way as for the theory of the constant
flow experiment. As in the previous paper, 3 it is supposed that
gas flows into the bubble, from a vessel at constant pressure,
through an orifice whose characteristic constant k is determined
by an experiment in which the gas under. consideration flows
steadily through the orifice in the absence of liquid. The orifice
is such that the volume flow rate of gas is k times the square
root of the pressure difference across the orifice, and when
the bubble is forming, the flow rate is given by:
JO
FLOW RATE (ml/s)
dV

dt
= k
[
P+ pgs
2a]
 r
t/2
(5)
Fig. 2.Experimenta/ and theoretical bubble volumes for large constant
flow rates
Fig. 3 shows bubble formation in groups of two. It will be
seen that the forming bubble is roughly spherical, and that
each bubble does not assume the spherical cap shape, characteristic of fully separated flow, 20 until it has moved some
distance from the orifice. These photographs therefore
ustify, to some extent, the assumption of unseparated flow
during formation.
Fig. 2 shows good agreement between theory and experin1ent
for flows between 15 and 20 ml/s, as in the range 1·53 ml/s
shown in Fig. 1. Data were not taken for the intermediate
Here P = P 1  pgh. Equations (1) and (5) were solved
numerically by the computer Edsac 2, with the initial conditions at t = 0:
ds
4
s = 0  = 0 Ji'.0 =  :n:r 3 •
' dt
'
3 °
The bubble was assumed to detach whens= r + r 0 , and the
results are plotted in the form of dimensionless groups in
Figs 4, 5, and 6. The dimensionless groups were formulated
from equations (1) and (5) with a and r 0 = 0, when there are
unique relations between the dimensionless bubble volume
V' = V/p 3 !4 k 3 ! 2 , the dimensionless pressure P' = P/gk 1 12p5 14,
Fig. 3.Bubble formation in groups af two with a constant flow rate of 13·7 ml/s. r 0 = 0·2 cm
TRANS. INSTN CHEM. ENGRS, Vol. 38, 1960
338
DAVIDSON AND scHi.iLER. BUBBLE FORMATION AT AN ORIFICE IN AN INVISCID LIQUID
the dimensionless flow G' = G/k 514g 112p518 • Here
V = V1  V 0 , and G is the mean flow rate during bubble
formation, and is J7 It x. where t 1 is the time at detachment. For
the case when a = 0 and r 0 = 0, these variables are related
and
• 0
s
2
0
Co1utant
prr1.su~c

>
• ·o
1~
l> ;'.l;Q,,,
>
\Con1lartt
J. 0
!low
v 1 : 1·37B {P') 315
V1
'.
•
p
/ _____ Constant flow
= 1·378 (G 1 )~
k
(g/ml) (dyo/cm) (<m) (crn71z 19 112)
' 0
I ·O
v
75·3
I ·0
•
1·1452
r · 1452
57.7
• 14
/ 5588
<48·2
0·16
2 ·036 0
0175
2 4356
0·20
3·1810
I 0
I 0
0
0·12
A
I
0
D
I 0
40· I
I· 0
29· 4
I 0
23 ·8
0225 4. 0 262
,.
I. 0
A
I 0
..
0 ·25
4·9706
Fig. 6.The theoretical relations between dimensionless bubble volume
p
(
and flow rate. At the point X, V'
• 0
3·.
2·0
'112
2
0
=
0·500, G'
= 0·351
)112
(P ) "" m5i4
•'
p
Fig. 4.The theoretical relations between dimensionless bubble volume
and pressure. When P' = 0, V' = 0·500 for constant pressure. The
points as well as the curves were derived from theory
•·•.
3·0
C onstGnt prcuur(
O ... r 0 • 0
by the curves shown in Figs 4, 5, and 6, and the diagrams also
show results for various values of a and r0 • It is apparent that
the effect of a and r 0 upon bubble volume and mean gas flow
rate is small. Surface tension docs however affect the minimum
value of P for bubbling to occur, since bubbling will cease if
Pis less than 2a/r 0 • It should also be noted that of Figs 4, 5,
and 6, only two are independent, and the third can be plotted
from the other two. In most practical problems the gas flow
rate is fixed beforehand, and the bubble volume V and the
pressure P have to be determined.
The significance of the diagrams can be better understood by
considering two limiting cases as follows:
k
Constant ft ow
G' = (P1)112
•
p
(glmi)(dyo/cm)
0
I 0
v
A
I ·1452
I 0
57·7
0 14
(·5566
, .0
48·2
0 ·16
2·0360
0
l·O
40· I
0·175
2 "4356
I 0
29·4
0·20
3·1810
l·O
23·8
0·225
4·0262
I 0
20·4
0·2S
4·9706
,..
H
'112 _ (
p)
1·1"452
0·12
p
f,
(
0
k
(cm7121 9 112)
753
I ·O
I0
'•
(cm)

3 0
p
1k112p514
••
)'"
Fig. 5.The theoretical relations between dimensionless flow rate and
pressure. When P' = 0, G' = 0·351 for constant pressure. The
points as well as the curves were derived from theory
SMALL; APPROACH TO CONSTANT FLOW
When the orifice constant k is very small, P will become
large for any flow to pass, and the flow will not vary during
bubble formation. Substituting G = kP 112 in equation (2) it
is easily shown that V' = 1378 (P') 3 / 5, G' = (P')1 12, and
these relations are shown plotted on Figs 4, 5, and 6, and, as
would be expected, they approach the "constant pressure"
curves aSy.mptotically.
In Fig. 5, the horizontal distance between the "constant
pressure" curve and the "constant flow" line is a measure
of the difference between the actual pressure drop through the
system and the pressure drop which would be obtained if the
flow through the orifice were uniform and equal to G. This
means that the normal method of calculating pressure drops
through bubbling devices by adding together a static liquid
head, a surface tension term, and a hole pressuredrop term
gives a total pressure drop which is too high, since all these
pressure drops do not act simultaneously.
TRANS. INSTN CHEM. ENGRS, Vol. 38, 1960
DAVIDSON AND SCHULER. BUBBLE FORMATION AT AN ORIFICE IN AN INVISCID LIQUID
p = 0;
TiiE CRITICAL FLOW
The "constant pressure" curves in Figs 4 and 5 show that
when P = 0, G and V have finite values. This discontinuity is
due to the presence in equation (5) of the term pgs, which
gives a finite pressure difference across the orifice as soon as
the bubble starts to move. With a finite value of the surface
tension, the minimum value of Pis 2a/r 0, and if Vis plotted as
a function of P, the curve will show a step up, at P = 2a/r0 ,
from zero to the value of V given by Fig. 4.
With a = 0 and P = 0, equation (5) becomes
dV/dt = k (pgs) 1! 2,
and this equation together with equation (I) has the following
analytical solution for r 0 = 0:
v=
8
gt·
33
2p)l/2
( 33
kgt2
s = 
0
(6)
339
very little, and this is in accordance with the theory which
predicts that V should stay constant when G is less than its
critical value. For this reason, the theoretical line on Fig. 7
is shown chain dotted, since when P = 968, j7 can be either
zero or 3·5. But in Fig. 8, the theoretical line is shown continuous, because at P = 968, all values of G between 0 and 67 are
possible. Intermediate values of G between 0 and 67 give a
bubble volume of 3·5 ml. Figs 7 and 8 show that when Pis
greater than 968 the theoretical flow rate and bubble volume
are considerably greater than the experimental values, and the
differences are thought to be due partly to the upward current
of water caused by the stream of rising bubbles, and partly
due to bubble deformation; these two factors were discussed
in connection with constant flow results shown in Fig. 2.
Fig. 9 shows bubble volume V plotted as a function of
mean flow G. When G is less than 67, the theoretical value of
The bubble will detach whens= (3 V/4n) 113 , giving the volume
V of the detached bubble and the mean flow rate Gas follows:
V'
=
j7
ks/2ps/4
=
G' = k5/4glf2p5/B
G
(33) 3/4
32
2112
=
(
.........
.,
3 ) 1/2
4n
= 0·500,
60
(7)
E
.___,.
(33)1/s (4n3)1/4 = 0·351.
4 2
(8)
These numbers agree well with the values calculated from
the computer results.
~
.,.........,........_,,.,.
0
.J
u.
40
_,.,.....
z
<(
::?
0
I
0
20
IC>
Experimental results with the "constant pressure" arrangement are shown in Table I, and Figs 7, 8, and 9 show the
results graphically for r0 = 0· 149 cm. These results show
clearly the phenomenon of the critical pressure, with a sudden
rise in bubble volume and mean flow rate when P Js_nearly
2a/r0 • Comparing Figs 7 and 8, it can be seen that when G
is reduced below its critical value (about 32 ml/s), V changes
_,.,.....
_,o
UJ
I
Experimental results
o_,.,.....
_,.,..... ......
I
0
900
I 000
I 300
I 200
1100
PPRESSURE {9 /cm s2 )
Fig. 8.Mean flow rate as a function of pressure, with air and water.
k = I ·90, r 0 = 0· 149 cm
T luory
4
4
Critical bubbl< volum<
/
E
~ 3
__s
::>
.J
UJ
>
3
x
,f
::?
0
w
al
3
0
>
2
o0
2
// f
w
CD
dl
.J
dl
I>
dl
::::>
CD
:::>
f
f
/
/
//
;~0·0222d
....,
1>
/
/
/
O'~'~'~~'~~~'~~~'~'
900
I 000
1100
P PRESSURE
I
200
I
300
(g /cm s 1 )
0
20
40
60
BO
G MEAN FLOW (m I /s)
Fig. 9.Bubble volume as a function of mean flow for constant pressure
Fig. 7.Bubble volume as a function of pressure with air and water.
k = 1·90, r 0 = 0·149 cm
TRANS. INSTN CHEM. ENGRS, Vol. 38, 1960
with air and water. k = I ·90, r 0
point of critical pressure
=
0· 149 cm.Xis the experimental
340
DAVIDSON AND SCHULER. BUBBLE FORMATION AT AN ORIFICE IN AN INVISCID LIQUID
a
e··.u
Fig. I0.Bubble formation at constant pressure with a mean flow rate below the critical value.
G = 26
ml /s, k
=
3·06 ml cm 1 ' 2/g 1 ' 2 , r o = 0· 187 cm
Fig. 11.Bubble formation at constant pressure with a mean flow rate above the critical value. G = 60 ml,'s, k = 3·06 ml cm 1 ' 2/g 1 i 2 , r 0 =0·187 cm
V is constant at 3·5, and the dotted curves are merely continuations of the theoretical curves for "constant pressure" and
"constant volume". At a given value of P, the experimental
values of V and G are both less than the theoretical, and consequently the points on Fig. 9 are shifted down towards the
bottom left hand comer of the diagram, and this shift includes
the critical fl.ow point X. In design work, G is the independent
variable, and Fig. 9 shows that for low values of G, the actual
value of V is less than the theoretical, but for high values of G
the simple formula from the "constant volume" theory gives
an approximate representation of the actual bubble volumes.
Table 1 shows that the characteristics of Figs 7, 8, and 9 are
accentuated for the larger orifices
In comparing "constant flow" and "constant pressure"
TRANS. INSTN CHEM. ENGRS, Vol. 38, 1960
DAVIDSON AND SCHULER. BUBBLE FORMATION AT AN ORIFICE IN AN INVISCID LIQUID
341
TABLE !.Formation of Air Bubbles in Water with Constant Gas Pressure
a = 72 dyn/cm
2a
p
k
(ml cm1t2/g1/2)
ro
(cm)
ro
(g/cm s2 )
(g/cm s2)
1·90
1 ·90
l ·90
1·90
1·90
3·06
3·06
3·06
3·06
3·82
3·82
3·82
3·82
4·90
4·90
4·90
4·90
0· 149
0· 149
0· 149
0· 149
0· 149
0· 187
0· 187
0· 187
0· 187
0·206
0·206
0·206
0·206
0·230
0·230
0·230
0·230
968
968
968
968
968
771
771
771
771
698
698
698
698
625
625
625
625
951
951
951
1118
1323
779
779
877
1024
734
734
832
1006
632
739
790
800
experiments (Figs 2 and 9), it should be noted that in the latter
experiments the effect of the "neck" formed during detachment
will be more important, because in the one case Vis proportional to t and in the other case to t 2 • This means that the
bubble formed in the "constant pressure" experiment will
receive a greater proportion of its gas during the detachment
period. Figs 10 and 11, which are cine pictures of bubbles
forming in a "constant pressure" experiment, do show the
very great rate of bubble expansion during detachment,
compared with the "constant flow" bubbles shown in Fig. 3.
Figs 10 and 11 also demonstrate bubble formation with
pressures above and below the critical. Fig. 10 shows that,
below the critical pressure, there is a discrete time interval
between one bubble and the nextthe magnitude of the
interval depending upon the size of the gas reservoir below the
orificewhereas Fig. 11 shows continuous bubble formation.
Conclusions
Experiments with a constant gas flow rate during bubble
formation have shown that equation (2) can be used as a
rough guide to the relation between the bubble volume V and
the gas flow rate G. This equation can also be used when the
bubbles form above an orifice with a constant supply pressure
below, provided the gas flow rate is somewhat above the
critical flow rate.
If the gas flow rate is less than or equal to this critical
flow rate, the bubble size will be roughly constant, and the
time interval between the bubbles will depend upon the volume
of the vessel below the orifice. If this vessel is infinitely large,
bubble formation will cease altogether when the pressure Pis
less than 2a/r 0 •
It seems likely that the critical flow rate described in the
present paper is closely related to the flow rate at which downward leakage or "dumping" will occur on a sieve plate,
because if the total gas flow divided by the number of holes is
less than the critical value of G, some of the holes will not be
bubbling continuously, and the opportunity for leakage ~ill
occur.
Acknowledgment
The authors wish to acknowledge the help of Mr. H.P. F.
SwinnertonDyer, who programmed the calculations for
Eclsac 2. One of us (J. F. D.) wishes to acknowledge the support of the University of Delaware during the time when this
paper was written.
TRANS. INSTN CHEM. ENGRS, Vol. 38, 1960
Mean flow G
Bubble volwne V
,A.,
........
Expt. Theory Expt. Theory
6/j
(ml/s) (ml/s)
(ml)
(ml)
0 · 0222 Gexpt,
17
23
32
45
61
26
33
47
60
17
30
57
68
25
60
68
70
67
67
67
70
76
102
102
105
112
124
124
129
141
156
163
169
169
2·0
2· 1
2·3
2·9
3·4
3·2
3·4
4· 1
4·5
4· 1
4·3
4·9
5·7
5·6
6·9
7· 1
7·5
3·5
3·5
3·5
3·8
4·2
6· 1
6· 1
6·4
6·9
7·8
7·8
8·3
9· l
10·7
11·4
11 ·7
11. 8
0·67
0·96
1 ·42
2· 14
3·06
1·11
1 ·48
2·26
3·04
0·67
l · 31
2·84
3·53
1·06
3·04
3·53
3·64
Symbols Used
G
=
constant gas flow rate.
G = mean flow rate in "constant pressure" experiment
G'
=
=Viti.
G/k•f4g1f2psfs.
= acceleration due to gravity.
h = depth of liquid seal.
k = orifice constant = (flow rate)/(pressure difference) 1 12 •
P =P1 pgh.
P 1 = pressure in the vessel below the orifice.
P' = P/g k1f2psf4.
r = radius of bubble at any instant.
r 0 = radius of orifice.
s = vertical distance moved by the bubble.
t
=time.
t 1 = time at detachment.
V = bubble volume at time t.
g
Vo = 4nr 0 3 /3.
Vi = bubble volume just before detachment.
V =Vi Vo.
V'
p
a
= V/p3/4ks12.
= liquid density.
= surface tension.
The above quantities may be expressed in any set of consistent units in which force and mass are not defined independently.
References
Mayfield, F. D., Church, W. L., Green, A. C., Lee, D. C. and
Rasmussen, R. W. Ind. Engng Chem., 1952, 44, 2238.
Brown, R. S. Ph.D. Thesis, Lawrence Radiation Laboratory,
_Berkeley, California, 1958.
8 Schiiler;l!. 0. G. and Davidson, J. F.
Trans. lnstn chem. Engrs,
1960, 38, 144.
4 Guyer, A. and Peterhans, F.
Helv. Chim. Acta., 1943, 26, 1099.
5 Datta, R. L., Napier, D. H., and Newitt, D. M.
Trans. Jnstn
chem. Engrs, 1950, 28, 14.
6 Van Krevelen, D. W. and Hoftijzer, P. J.
Chem. Engng Progr.,
1950, 46, 29.
7 Coppock, P. D. and Meiklejohn, G. T.
Trans. lnstn chem. Engrs,
1
2
1951, 29, 75.
Benzing, R. J. and Myers, J. E. lndustr. Engng Chem., 1955,
47, 2087.
• Eversole, W. G., Wagner, G. H. and Stackhouse, E. lndustr.
Engn Chem., 1941, 33, 1459.
1° Calderbank, P. H. Trans. lnstn chem. Engrs, 1956, 34, 70
8
342
11
DAVIDSON AND SCHULER. BUBBLE FORMATION AT AN ORIFICE IN AN INVISCID LIQUID
Siemes, W. and Kaufmann, J. F.
Chemical Engineering Science,
1956, 5, 127.
Hughes, R.R., Handlos, A. M., Evans, H. D. and Maycock, R. L.
Publication of the Heat Transfer and Fluid Mechanics
Insitute, Los Angeles, 1953.
13 Quigley, C. J., Johnson, A. I. and Harris, B. L.
Chemical
Engineering Progress Symposium Series No. 16, 1955, 31.
14 Davidson, L. and Amick, E. H.
A.I.Ch.£. Journal, 1956, 2, 337.
15 Leibson, I., Holcomb, E. G., Cacoso, A. G. and Jacmic, J. J.
A.I.Ch.£. Journal, 1956, 2, 296.
16 Hayes, W. B., Hardy, D. W., and Holland, C. D.
A.I.Ch.E.
Journal, 1959, 5, 319.
12
Prandtl, L. and Tietjens, 0. G. "Applied Hydro and Aeromechanics'', 1934, (New York: McGrawHill Book Co.).
18 Goldstein, S., "Modern Developments in Fluid Dynamics,"
1938 (Oxford: The University Press).
19 Helsby, F. W. and Tuson, K. R.
Research, 1955, 8, 270..
20 Davies, R. M. and Taylor, G. I.
Proc. Roy. Soc., 1950, A200,
17
375.
The manuscript of this paper was received on 26 May, 1960 and the
paper was presented at a symposium of the Institution in London on
10 October, 1960.
TRANS. INSTN CHEM. ENGRS, Vol, 38, 1960
Link to this page
Permanent link
Use the permanent link to the download page to share your document on Facebook, Twitter, LinkedIn, or directly with a contact by eMail, Messenger, Whatsapp, Line..
Short link
Use the short link to share your document on Twitter or by text message (SMS)
HTML Code
Copy the following HTML code to share your document on a Website or Blog