Вы находитесь на странице: 1из 11

Chemical Engineering Science 56 (2001) 53715381

www.elsevier.com/locate/ces

Aperiodic bubble formation from a submerged ori&ce


Lei Zhang , Masahiro Shoji
Department of Mechanical Engineering, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan

Received 9 January 2001; received in revised form 24 May 2001; accepted 28 May 2001

Abstract
The present work aims at studying the nonlinear behaviors of bubbles, including formation, interference, collision and coalescence,
formed from a submerged ori&ce. The experimental data reveal that the departing periods of successive bubbles evolve regularly
from single period to triple periods within the air 5ow rate regime of 100 cc=min q 2000 cc=min, due to the interactions between
the consecutive bubbles. A new comprehensive theoretical model is developed for describing the instantaneous bubble behaviors
during formation and ascendance processes and for predicting the departing periods and sizes of successive bubbles in constant 5ow
rate condition. Owing to the estimation of instantaneous interactions between successive bubbles and the incorporation of the wake
e8ect of previous bubble, the present model can elaborately describe the evolution process and mechanisms of bubble departing
periods corresponding to di8erent gas 5ow rate regimes. The theoretical results are in good agreement with the experimental
investigation. Both the experimental data and the theoretical results state that bifurcation of bubble departing periods in constant
5ow rate condition is induced by bubble interactions. ? 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Bubble behavior; Aperiodic bubble departure; Bubble formation; Bubble interaction; Nonlinear model

1. Introduction Liow & Gray, 1988; Terasaka & Tsuge, 1990, 1993; Oguz
& Prosperetti, 1993).
The behaviors of bubbles, especially their formation, On the other hand, a great number of experiments have
interaction and coalescence, are of key importance in been directed at the study of bubbling regimes in the
such &elds as gasliquid contactor, gasliquid separator, constant 5ow rate condition (McCann & Prince, 1971;
bubble absorption, boiling, bubble column, fermentation, Krevelen & Hoftijzer, 1950; Wraith, 1971; Kyriakides,
cavitations, etc. Despite the numerous theoretical and ex- Kastrinakis, & Nychas, 1997). It is found from the ex-
perimental investigations of single bubble formation, the periments that except for the low gas 5ow rate regime,
mechanism of various bubble behaviors at submerged ori- the departing period and size of a bubble are in5uenced
&ces remains far from fully understood. by the wake of the previous bubbles. Referring to the
In the past several decades, pioneered by David- considerable experimental investigation of bubble inter-
son (Davidson & Schuler, 1960a, b), the single bubble actions in a bubble column, the drag force of the fol-
model has been developed to predict the single bubble lowing bubble is released within the wake of the previ-
formation process in viscous and inviscid liquid more ous bubble and then the following bubble is accelerated.
accurately. However, almost all the theoretical models This acceleration e8ect &nally evokes the coalescence of
can only simulate single bubble formation while the these two bubbles and the coalescence position becomes
interactions of successive bubbles have been neglected closer to the ori&ce exit with the increase of gas 5ow rate
(Ramakrishnan, Kumar, & Kuloor, 1969; Marmur & (Leibson, Holcomb, Cacoso, & Jasmic, 1956; Kupfer-
Rubin, 1976; Princzewski, 1981; Tan & Harris, 1986; berg & Jameson, 1969; Narayanan, Goossens, & Kossen,
1974; Acharya & Ulbrecht, 1978; Bhaga & Weber, 1980;
Corresponding author. Fax: +81-3-5841-6408. Miyahara, Haga, & Takahashi, 1983; Dekee, Carreau, &
E-mail address: zhang@photon.t.u-tokyo.ac.jp (L. Zhang). Mordarski, 1986; Li, 1999).

0009-2509/01/$ - see front matter ? 2001 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 0 9 - 2 5 0 9 ( 0 1 ) 0 0 2 4 1 - X
5372 L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381

Therefore, it can be noted that the bubble theoretical


models, which do not take into account the bubble inter-
actions, can only be limitedly applied to the low 5ow rate
regime, while at higher gas 5ow rate, the predicted results
expectedly deviate from the experimental data. Although
Chuang (Chuang & Goldschmidt, 1970) and Deshpande
(Deshpande, Deo, Hanson, & Oblad, 1992) improved the
model of Ramakrishnan (Ramakrishnan et al., 1969) by
taking into account the wake e8ect of the previous bubble
or including bubble coalescence in the model, the evolu-
tion of nonlinear bubble behaviors and the resulting e8ect
on bubble size and departing period cannot be predicted
exhaustively for di8erent gas 5ow rate regimes. These
problems are the topics of consideration in the present
paper. Fig. 1. Schematic diagram of experiment apparatus.
In the present research, a series of experiments are &rst
carried out to investigate the relation between the bubble
behaviors and the aperiodic characteristics of bubble de- surrounding the bubbles and the results are analyzed in
tachment at a submerged ori&ce within a wide air 5ow another paper (Abe, Xu, Kojima, Chin, & Shoij, 2000).
rate range. A new nonlinear bubble model is provided,
which can be used to estimate the instantaneous bubble 2.2. Results and analysis
behaviors and accurately predict the bubble size and de-
parting period within di8erent gas 5ow rate regimes. The The bubble departing period  is de&ned as the time
theoretical results agree with the experimental data satis- interval between the departing moment of a bubble and
factorily. that of the next bubble. The bubble departing moment
is identi&ed when the bubble neck becomes zero. From
the bubble departing period and the air 5ow rate, the
2. Experimental research bubble size distribution can be estimated. The bub-
bling phenomena above a submerged ori&ce can be
A series of experiments have been carried out to study divided into four regimes within the air 5ow rate range
the bubble behaviors at a submerged ori&ce in the air 100 cc=min q 2000 cc=min. These four regimes are
water system (Abe, Xu, Kojima, Chin, & Shoij, 2000). single bubbling, pairing, double coalescence and triple
The interest was aroused by the fact that the sizes and bubble formation (shown in Fig. 2). In di8erent regimes,
the departing periods of successive bubbles are perfectly the aperiodic characteristics of bubble departing intervals
uniform at a low air 5ow rate, while with the increase of are distinct.
air 5ow rate, the aperiodic characteristics of bubble de- Single bubbling regime extends up to an air 5ow rate
parture gradually become distinct in spite of the constant of about 300 cc=min (see Fig. 2(a)). In this regime, the
5ow rate condition. Furthermore, the evolution of bubble spherical bubbles are formed after every strictly identical
departing periods is regular within the air 5ow rate range time interval at a certain air 5ow rate. During the forma-
of 100 cc=min q 2000 cc=min in the experiments. tion process of a bubble, the in5uence from the wake of
the previous bubble seems negligible.
2.1. Experiment apparatus Pairing regime (shown in Fig. 2(b)) occurs in the air
5ow rate regime from 300 to 700 cc=min. This regime
An overview of the experiment apparatus is shown in is characterized by the fact that during the formation
Fig. 1. The air supplied by the compressor &rst 5owed process, the bubble elongates in the vertical direction
through a long capillary tube and a porous medium, owing to the in5uence of the wake of the previous
before it reached the ori&ce submerged in a stagnant bubble. Consequently, the formation process of this fol-
water pool, in order to set up the constant 5ow rate lowing bubble is accelerated slightly compared to the
condition. The diameter of the ori&ce investigated was previous one (1 = 33 ms; 2 = 30 ms). In the lower part
2 mm. The air volume 5ow rate is regulated from 100 to of the 5ow rate region (300 cc=min q 450 cc=min),
2000 cc=min. The cross section of the tank was 20 cm the following bubble only interferes with the pre-
20 cm and the tank was &lled with distilled water to a vious bubble but does not collide with it during
height of 18 cm. For the visual observations, a high-speed the formation process. While for the upper part re-
video camera was used to record the various bubble be- gion (450 cc=min q 700 cc=min), the following
haviors in the formation and ascendance processes. A hot bubble collides with the previous bubble without
&lm probe was used to study the liquid micro-convection coalescence.
L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381 5373

Fig. 3. Evolution of bubble departing periods with the increase of air


5ow rate.

arise. In this case, the size of the previous bubble is much


larger than that of the following one. The &rst double
coalescence process is over at t = 53 ms, then the sec-
ond double coalescence occurs between t = 53 ms and
t = 106 ms repeatedly.
The pairing of an already formed double bub-
ble with the consecutive single bubble is termed
here as triple bubble formation. In this regime (see
Fig. 2(d)), with the increase of the air 5ow rate
(1400 cc=min q 2000 cc=min), the double coales-
cence occurs earlier than that in the double coalescence
regime. Therefore, the following bubble is absorbed
by the previous bubble in a even shorter time interval
(3 = 13 ms). In this case, the size of the previous bubble
generally is 2 or 3 times that of the following bubble.
Furthermore, because the successive third bubble grows
very quickly in this air 5ow rate regime, it collides
with the former double bubbles violently and deforms
remarkably. Due to the wake 5ow in5uence of the former
Fig. 2. (a) Single bubbling (q = 100 cc=min; ReO = 68); (b) double bubbles, the third bubble detaches from the ori-
pairing (q = 350 cc=min; ReO = 238); (c) double coalescence
(q = 900 cc=min; ReO = 612); and (d) triple bubble formation
&ce faster than the &rst bubble (1 = 35 ms; 3 = 27 ms).
(q = 1500 cc=min; ReO = 1020). In succession, another triple bubble formation takes
place between t = 74 ms and t = 148 ms repeatedly with
similar mechanism.
When the air 5ow rate increases to a certain value From the experiments, it is found that the bubble de-
(q = 700 cc=min), the two successive bubbles coalesce parting periods evolve regularly from single period to
during the formation process of the following one. This triple periods with the increase of the air 5ow rate from
is the symbol of double coalescence regime. As shown in 100 to 2000 cc=min. The relation between the air 5ow
Fig. 2(c), from t = 0 ms to t = 33 ms, the previous bub- rate and the bubble departing periods in the above four
ble forms and detaches from the ori&ce without interfer- regimes is shown in Fig. 3. (the pairing regime is named
ence with its former bubbles. From t = 33 ms, the follow- single period region together with the single bubbling
ing bubble begins to grow and then coalesces with the regime).
previous bubble during the formation process. As coa-
lescence occurs, the following bubble elongates remark-
ably in the vertical direction and then seems to be pulled 3. Theoretical model
away from the ori&ce and absorbed by the previous bub-
ble within a short time interval (2 = 20 ms). The evident An exact theoretical approach to describe the bubble
bifurcation phenomena of the bubble departing period and surrounding liquid motion before and after bubble
5374 L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381

detachment is highly complicated. While within the rela- Combining the two kinds of drag force to one term,
tive low gas 5ow rate range, the micro-convection e8ect the overall drag force exerted on the forming bubble can
of liquid is not strong, the most signi&cant in5uence upon be described as
the forming bubble is from the wake e8ect of its previ-   
1  2 ds  ds 
ous bubble (Nogami, Ogusu & Shoij, 2001). Therefore,

Fd = l D CD 
vw  vw  : (9)
2 4 dt dt
based on certain simpli&ed assumptions, the nonlinear
theoretical model developed in this section mainly takes From Eq. (9), we can &nd that the entire drag force has
into account the wake e8ect of the former bubbles, and in dual e8ects and can act as an aiding or an inhibiting
particular the resulting various interactions between suc- force, depending on the relative velocity of the forming
cessive bubbles, including interference, collision and co- bubble with respect to the surrounding wake induced by
alescence. This model can estimate the bubble behaviors the previous bubble.
and predict the bubble departing periods and sizes within In the present model, bubble formation is assumed to
relatively low gas 5ow rate regimes. take place in two stagesthe expansion stage and the
elongation stage. During the expansion stage, the bubble
3.1. Bubble formation grows radially due to the incoming gas 5ux, but the bub-
ble base remains attached to the ori&ce. In this case, the
The bubble shape is assumed to be spherical through- bubble center velocity ds=dt is equal to the bubble radial
out the formation process. Thus, under the constant 5ow expansion velocity dr=dt, that is
condition, the variation rate of bubble volume is given as ds dr
= : (10)
dV dt dt
= q = const: (1)
dt This stage ends when the sum of detaching forces is
The present model is based on an overall force balance larger than or equal to that of attaching forces, which can
acting on the bubble (shown in Fig. 5(a). The theoretical be stated as
description of every force is stated as follows:
FB + FM F + Fi + Fd : (11)
Buoyancy force: FB = V (l g )g: (2)
During the elongation phase, additional gas is fed into
q2 the bubble, which continues to grow in size. The bubble
Gas momentum 7ux: F M = g : (3)
(=4)DO2 lifts away from the ori&ce, but is still attached to it through
a neck, which also grows with time. During the elongation
Surface tension force: F = DO : (4) stage, the force balance always remains as
FB + FM = F + Fi + Fd : (12)
Added mass inertia force: In this stage,
  ds dr dl
d ds = + ; (13)
Fi = (g + 1 l ) V ; (5) dt dt dt
dt dt
where 1 is the added mass coeOcient de&ned as 11=16 where dr=dt is the bubble radial expansion velocity and
(Lamb, 1932). dl=dt is the bubble neck elongation velocity.
 2 At the end of the second stage, the neck pinches o8
1  ds and the bubble detaches. Following the detachment cri-
Drag force: Fd = l D2 CD ; (6)
2 4 dt terion discussed by Kim, Kamotani, and Ostrach (1994),
where CD = 18:5=ReB0:6 for 1 ReB 1000 and CD = 0:44 it is assumed that the bubble neck collapses when the
for ReB 1000 (Sekoguchi, 1974). neck length becomes larger than or equal to the ori&ce
diameter,
Drag force due to the wake of the former bubble:
l = S r DO : (14)
1 
Fdw = l D2 CD (vw )2 ; (7)
2 4
3.2. Bubble detachment and ascendance
where vw is the average impressed velocity due to the
wake of the previous detached bubble. Schlichting (1960) At the moment of detachment, because the bubble neck
showed that the velocity defect behind a circular wake is cut o8 suddenly, the surface tension force and the gas
is proportional to the 2=3 power of the axial distance. momentum 5ux disappear immediately at the same time.
Extending this to the two successive bubbles, it can be The force balance of the bubble is destroyed. But due to
expected that the inertia, the upper part of the detached bubble keeps
 
Sp hp 2=3 its shape and position immediately after detachment,
vw = vp : (8)
hp Sf while the lower part shrinks towards the bubble center
L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381 5375

Fig. 4. Comparison of bubble departing, ascendance and deformation processes within di8erent air 5ow rate regimes: (a) q = 100 cc=min; (b)
q = 800 cc=min.

(see Fig. 4(a) and (b) t = 0 ms; t = 1:5 ms). Furthermore, calculate the deform coeOcient , instantaneously during
when the gas 5ow rate is low, the following bubble does the simulation process.
not begin to grow immediately after the previous bubble After detachment, the bubble reaches a new force bal-
detached. A short waiting time has to be considered ance by interactions with the surrounding liquid. The
between the detachment of the previous bubble and forces exerted on the detached bubble are assumed to in-
the formation of the following bubble (shown in clude buoyancy force FB , inertia force Fi and drag force.
Fig. 4(a) t = 0 ms; t = 6:0 ms; S3 S1 ), while with the The drag force due to the wake of the previous bubble is
increase of the gas 5ow rate, the next bubble begins to neglected during the bubble ascent process. Based on the
grow immediately after the previous bubble detaches new force balance (shown in Fig. 5(b))
(shown in Fig. 4(b) t = 0 ms; t = 1:5 ms; S3 = S1 ). It
FB = Fi + Fd ; (16)
should be noted that the minimum time interval of
high-speed video camera is 1:5 ms. the bubble ascent velocity v can be calculated.
In order to consider the variation of detachment dis-
tance with the increase of gas 5ow rate, the following 3.3. Bubble interference
assumption is employed:
In the single bubbling regime, the distance between the
hd = (1 + +)Se ,re ; (15) two successive bubbles is large and the wake velocity of
the previous bubble vw is negligibly small (see Fig. 6(a)).
where for single bubbling regime, + = 0:1; , = 0:2 and The criterion is
for other regimes, + = 0; , = 0:2, mainly based on the
video observation. But the deforming coeOcient , is only hp Sf rf (1 ,)rp : (17)
an average value. Since, as shown in Fig. 4, during the However, when the gas 5ow rate increases up into
ascendance process of the previous bubble, the bubble the pairing regime, the distance between the two suc-
deforms violently and continuously, it is very complex to cessive bubbles is so short that Fdw has to be taken
5376 L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381

Fig. 5. Schematic diagram of the end of elongation stage (a), bubble departing moment (b), and bubble ascendance (c).

Fig. 6. Evolution of bubble interactions: (a) without interference, (b) bubble interference, (c) bubble collision, and (d) bubble coalescence.

into account (see Fig. 6(b)). The criterion of interference where 2 is given as 1.8 in the present model based on
is assumed as the experimental data.

0 hp Sf rf (1 ,)rp : (18) 3.4. Bubble collision


The interference region is shown in Fig. 5(b) as shadow
In the upper part of the pairing regime, the following
area and the above positional variables are shown in
bubble contacts and collides with the previous bubble
Fig. 5(c). When the upper surface of the following bub-
during formation process but without coalescence. This is
ble enters this region, the interference with the previous
a transition stage between bubble interference and bubble
bubble occurs. Otherwise, the following bubble forms
coalescence (shown in Fig. 6(c)). The criterion of bubble
without interference.
collision is assumed to be
Furthermore, when the interference occurs, the fol-
lowing bubble deforms in the vertical direction due hp Sf rf 6 0; (20)
to the wake e8ect of the previous bubble. It is as-  
ds
sumed that when the interference occurs, the shape of (g + 1 l )Vf vp Qt Dp : (21)
dt
the following bubble changes from sphere to ellipsoid.
The two horizontal radii of the ellipsoid are identi- Eq. (20) states that the distance between the lower sur-
cal and de&ned as rh . The vertical radius is assumed face of the previous bubble and the upper surface of the
to be following bubble is less than or equal to zero. In Eq. (21),
which is deduced from the momentum balance, the right
rv = 2 rh ; (19) term is the surface tension force of the previous bubble
L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381 5377

and the left term states the collision force, which is in- After coalescence, the two bubbles reach a common
duced by the momentum decrease of the following bub- velocity vc , which is taken as the initial velocity for the
ble during the collision time Qt. It is assumed that within elongation process of the two coalesced bubbles, by solv-
the collision time, Qt, the center velocity of the follow- ing the momentum conversation equation as follows:
ing bubble ds=dt decreases to the rising velocity of the  
ds
previous bubble vp . The collision time Qt is de&ned as (g + 1 l )Vp vp + (g + 1 l )Vf
dt
1
Qt = ; (22) = (g + 1 l )(Vp + Vf )vc ; (26)
3!

where 3 is a constant, ! is the single bubble oscillation where ds=dt can be obtained by the force balance equation
period de&ned as (12) after coalescence. During the following elongation
 process, the double coalesced bubbles elongate together
2
!= : (23) at the velocity of (vc + ds =dt), that is
 (g + 0:5l )rp3
2
ds ds
= vc + ; (27)
Eq. (21) states that when the collision force is less than dt dt
the surface tension force of the previous bubble, the thin where ds =dt is the center velocity of the double bubbles
liquid &lm, which separates the two bubbles, cannot be due to the accumulation of the gas 5ow rate fed by the
penetrated. So these two bubbles only collide without ori&ce.
coalescence. Bubble coalescence in elongation stage. While coa-
In addition, when the following bubble collides with lescence occurs during the elongation stage of the fol-
the previous bubble, the following bubble also deforms lowing bubble, it is assumed that the following bubble
from sphere to ellipsoid, which is similar to that in the does not deform again but reaches the common velocity
bubble interference condition. vc , which can be obtained by Eq. (26), due to the inter-
actions between these successive bubbles. The following
3.5. Bubble coalescence elongation process is similar to that in the previous sec-
tion of bubble coalescence in expansion stage.
When the gas 5ow rate increases up to the double coa- After detachment, these two coalesced bubbles ascend
lescence regime, the center velocity of the following bub- upwards as an entity, which is named double bubble. Then
ble during the formation process consequently increases this double bubble also interferes, collides or coalesces
so rapidly that the resulting collision force becomes larger with the following bubbles. Therefore, with the increase
than the surface tension force of the previous bubble. In of gas 5ow rate, the interactions between successive bub-
this case, the thin liquid &lm is penetrated and then the bles become more and more complicated.
two bubbles coalesce (shown in Fig. 6(d)). Therefore, In the above model, two constants 3 and 4 remain un-
the criterion for bubble coalescence can be described as known. In the numerical simulation, the values of 3 = 20
hp Sf rf 6 0; and 4 = 4:0 are employed to best &t to the experimental
  (24) data.
ds
(g + 1 l )Vf vp Qt Dp :
dt
From experiments, it is found that after coalescence, the 4. Numerical sketch and simulation results
following bubble elongates remarkably in the vertical di-
rection and the bubble neck immediately begins to grow. 4.1. Numerical sketch
Based on this discovery, the bubble coalescence is further
This model aims to study the bubble forming behav-
divided into two conditions: bubble coalescence in ex-
ior, especially its departing periods. From experiments it
pansion stage and bubble coalescence in elongation stage.
is found that the most signi&cant interaction exists only
Bubble coalescence in expansion stage. While the two
between two successive bubbles. However, if the form-
bubbles coalesce in the expansion stage of the following
ing bubble coalesces with its previous bubble, then the
bubble, the following bubble is assumed to deform again
behavior of this double bubble is also in5uenced by its
rapidly (the following bubble has deformed so as to be
previous bubble. Therefore, in the simulation process, the
ellipsoid when the bubble interference occurs). After de-
interactions among only three bubbles, the present form-
formation it reaches the force balance; consequently, ex-
ing bubble and its two immediate previous bubbles, are
pansion stage ends and elongation stage begins. The re-
taken into account simultaneously. The simulation pro-
quired time for this process is de&ned as coalescence time
cedure is simply demonstrated in Fig. 7(a) and (b).
1 During the simulation process of a bubble formation,
Qtc = ; (25)
4! in every time step (h = 0:0001 s), the interaction types
where 4 is a constant. between this forming bubble and its previous bubble
5378 L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381

Fig. 7. (a) Main calculation 5ow diagram; (b) 5ow diagram of estimate interaction types between the following bubble and the previous bubble.
L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381 5379

Fig. 7. Continued.

are &rst estimated, including interference, collision and


coalescence. Next the processes of expansion, elongation,
detachment and ascendance of this bubble are simulated
in succession. Finally, the departing period, the size and
the ascendance velocity of this bubble can be obtained
and saved for the calculation of the next bubble. Consec-
utively, the formation and ascendance processes of suc-
cessive following bubbles can be simulated by a similar
procedure one by one.
In addition, all the di8erential equations are solved by
the fourth-order Rungakutta method.

4.2. Results and discussion

Based on the above theory, the formation and as- Fig. 8. Comparison of predicted dimensionless bubble volume with
cendance processes of successive bubbles are simu- experimental data (Abe, Xu, Kojima, Chin, & Shoij, 2000).
lated within a relatively large gas 5ow rate regime
from 100 to 2500 cc=min. The comparison of theoret-
ical dimensionless bubble volumes with experimental regime, the bifurcation of bubble departing period is
data is stated in Fig. 8, where dimensionless volume induced by the coalescence of two successive bub-
is the ratio of bubble volume to the basic bubble vol- bles. In the triple bubble formation regime, the sec-
ume V0 , which is obtained when the buoyancy force ond bubble coalesces with the &rst bubble and the
is assumed to be equal to the surface tension force. third bubble interferes with the previous coalesced
It can be found that the present model prediction and double bubbles. In consequence, the triple periods
the experimental results are in good agreement. In the arise. Therefore, bubble interaction is the key rea-
single bubbling regime, every bubble forms without son for the variation of bubble departing periods and
interference with the previous bubbles, so the depart- sizes in constant 5ow rate condition, which can be
ing periods are always uniform. In the pairing regime, discovered by the experiments and revealed by the
interference between the following and the previous present model.
bubbles results in the slight distinction of successive The comparison of Krevelens experimental results
bubble departing periods. In the double coalescence (Krevelen & Hoftijzer, 1950) with the present theoretical
5380 L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381

experimental data has been achieved. The experimental


results and the theoretical prediction both demonstrate
that bubble interaction is the primary reason for the
aperiodic characteristics of bubble departure.
However, with the increase of gas 5ow rate, the
micro-convection e8ect of liquid becomes signi&cant
and should not be neglected, so strictly speaking, the
present model can only be applicable to the relatively
low gas 5ow rate regime. Moreover, the in5uence of
the ori&ce is very complex and needs further research in
future.

Notation
Fig. 9. Comparison of predicted dimensionless bubble volume and
experimental data (Krevelen et al., 1950).
CD drag force coeOcient
D bubble diameter
DO ori&ce diameter
FB buoyancy force
results is shown in Fig. 9. In this &gure, the predicted
FM gas momentum 5ux
bubble volumes in the coalescence regimes, such as
F surface tension force
double coalescence and triple bubble formation regimes,
Fi inertia force
represent the overall volumes of the coalesced bubbles. It
Fd drag force
can be found that within the coalescence region, almost
Fdw drag force due the wake of the previous bubble
all the measured data of average bubble volumes dis-
Fd overall drag force
tribute between two curves, which denote the predicted
g gravitational acceleration
values of single bubble volume and that of coalesced
h distance from lower surface of bubble
bubble volume, respectively. Consequently, after coales-
to ori&ce
cence occurs, the volumes of successive bubbles actually
l bubble neck length
are not identical, which has not been revealed by the dl
bubble neck elongation velocity
experimental data. dt
q gas 5ow rate
Therefore, all the single bubble models, in which
r bubble radius
the bubble interactions have not been taken into
ReB bubble Reynolds number, ReB = (D (dr=dt))=/l
account, cannot correctly predict the bubble size
ReO ori&ce Reynolds number, ReO = ug DO =/g
and explain the evolution of bubble departing pe- dr
riod, especially during the coalescence region. How- dt bubble radial expansion velocity
ever, all these targets can be achieved through the S distance from bubble center to ori&ce
present model. ds=dt bubble center velocity
t time from start of bubble expansion
Qt collision time
Qtc coalescence time in the expansion stage
5. Conclusions ug gas injecting velocity, ug = 4q=DO2
V bubble volume
A comprehensive study is developed on the various be- V0 basic bubble volume, V0 = DO =(l g )g
haviors of bubbles formed from a submerged ori&ce under v ascendance velocity of detached bubble
constant 5ow rate condition. The experimental investi- vc bubble velocity after coalescence
gation reveals that the departing periods of successive vw average impressed velocity of the previous
bubbles evolve regularly from single period to triple peri- bubble wake
ods within the four di8erent bubbling regimes, due to the
interactions between the consecutive bubbles. The new Greek letters
theoretical model provided can elaborately describe
the evolution mechanisms of bubble departing periods , bubble shrink coeOcient after detachment
and exactly predict the bubble size with the increase l liquid density
of gas 5ow rate, owing to the estimation of instan- g gas density
taneous interactions between successive bubbles and  surface tension
the incorporation of the wake e8ect of the previous  bubble departing period
bubble. Good agreement of the theoretical results with the ! single bubble oscillation period
L. Zhang, M. Shoji / Chemical Engineering Science 56 (2001) 53715381 5381

Subscripts Kyriakides, N. K., Kastrinakis, E. G., & Nychas, S. G. (1997).


Bubbling from nozzles submerged in water: transitions between
c coalescence bubbling regimes. Canadian Journal of Chemical Engineering,
75, 684691.
d departing moment
Lamb, H. (1932). Hydrodynamics (pp. 130 133). New York: Dover.
e end of elongation stage Leibson, I., Holcomb, E., Cacoso, A., & Jasmic, J. (1956). Rate of
f following bubble 5ow and mechanics of bubble formation from single submerged
p previous bubble ori&ces. A.I.Ch.E. Journal, 2, 296306.
Li, H. Z. (1999). Bubbles in non-newtonian 5uids: Formation,
interactions and coalescence. Chemical Engineering Science, 54,
References 22472254.
Liow, J. L., & Gray, N. B. (1988). A model of bubble growth in
Abe, N., Xu, L., Kojima, R., Chin, S. Y., & Shoij, M. (2000). wetting and non-wetting liquids. Chemical Engineering Science,
Experimental research on nonlinear characteristics of bubble 43, 31293139.
formation and interference. Proceedings of the 37th national heat Marmur, A., & Rubin, E. (1976). A theoretical model for bubble
transfer symposium of JAPAN, vol. 2, (pp. 435 436). formation at an ori&ce submerged in an inviscid liquid. Chemical
Acharya, A., & Ulbrecht, J. J. (1978). Note on the in5uence of Engineering Science, 31, 453463.
viscoelasticity on the coalescence rate of bubbles and drops. Miyahara, I., Haga, N., & Takahashi, T. (1983). Bubble formation
A.I.Ch.E. Journal, 24, 348351. from an ori&ce. International Chemical Engineering, 23, 524531.
Bhaga, D., & Weber, M. E. (1980). In-line interaction of a pair of McCann, D. J., & Prince, R. G. H. (1971). Regimes of bubbling at a
bubbles in a viscous liquid. Chemical Engineering Science, 35, submerged ori&ce. Chemical Engineering Science, 26, 15051512.
24672474. Narayanan, S., Goossens, L. H. J., & Kossen, N. W. F. (1974).
Chuang, S. C., & Goldschmidt, V. W. (1970). Bubble formation Coalescence of two bubbles rising in line at low Reynolds numbers.
due to a submerged capillary tube in quiescent and co5owing Chemical Engineering Science, 29, 20712082.
streams. American Society of Mechanical Engineers, 70-FE-21, Nogami, S., Ogusu, T., & Shoij, M. (2001). Nonlinear interaction
17. between bubble generation and micro-convection. Master thesis,
Davidson, J. F., & Schuler, B. O. (1960a). Bubble formation at University of Tokyo.
an ori&ce in a viscous liquid. Transactions of the Institution of Oguz, H. N., & Prosperetti, A. (1993). Dynamics of bubble growth
Chemical Engineers, 38, 144154. and detachment from a needle. Journal of Fluid Mechanics, 257,
Davidson, J. F., & Schuler, B. O. (1960b). Bubble formation at an 111145.
ori&ce in an inviscid liquid. Transactions of the Institution of Princzewski, W. V. (1981). The formation and growth of bubbles at
Chemical Engineers, 38, 335342. a submerged ori&ce. Chemical Engineering Science, 36, 405411.
Dekee, D., Carreau, P. J., & Mordarski, J. (1986). Bubble velocity and Ramakrishnan, S., Kumar, R., & Kuloor, N. R. (1969). Studies
coalescence in viscoelastic liquids. Chemical Engineering Science, in bubble formationI: Bubble formation under constant 5ow
41, 22732283. conditions. Chemical Engineering Science, 24, 731747.
Deshpande, D. A., Deo, M. D., Hanson, F. V., & Oblad, A. G. (1992). Schlichting, H. (1960). Boundary layer theory. New York:
A model for the prediction of bubble size at a single ori&ce in McGraw-Hill.
two-phase gasliquid systems. Chemical Engineering Science, 47, Sekoguchi, K. (1974). Kikai no kenkyu, vol. 1, Two-phase 5ow
16691676. (pp. 234 252). Tokyo: Yokendo Ltd.
Kim, I., Kamotani, Y., & Ostrach, S. (1994). Modeling bubble Tan, R. B. H., & Harris, I. J. (1986). A model for non-spherical
and drop formation in 5owing liquid in microgravity. A.I.Ch.E. bubble growth at a single ori&ce. Chemical Engineering Science,
Journal, 40, 1928. 41, 31753182.
Krevelen, D. W. van., & Hoftijzer, P. J. (1950). Studies of gas-bubble Terasaka, K., & Tsuge, H. (1990). Bubble formation at a single
formationcalculation of interfacial area in bubble contractors. ori&ce in highly viscous liquids. Journal of Chemical Engineering
Chemical Engineering Progress, 46, 2937. of Japan, 23, 160165.
Kupferberg, A., & Jameson, G. J. (1969). Bubble formation at Terasaka, K., & Tsuge, H. (1993). Bubble formation under constant
a submerged ori&ce above a gas chamber of &nite volume. 5ow conditions. Chemical Engineering Science, 48, 34173422.
Transactions of the Institution of Chemical Engineers, 47, Wraith, A. E. (1971). Two stage bubble growth at a submerged plate
241250. ori&ce. Chemical Engineering Science, 26, 2635.

Вам также может понравиться