References

Becker, R.. Ann. Physik. 32, 128 (1938).

Becker, R.. Dnring, Lf-., Zbid.. 24, 719 (1935).
Bobalek, E. G.. Off.Digest 34, 1293 (1962).
Bradley, R . S.,Quart. Xer.. (London) 5 , 315 (1951).
Cheever. G. D., Bobalek. E. G.. Of. Dicest 35, 759 [1963),
Cheever, G. D.. Parker. M. T., Rohileky E. G:. Ibtd, 34, 1047
1962).
(7) Chekick. J. J., J . Phys. Chem. 66, 762 (1962).
(8) Cook, H. D.. Ries. H. E., Jr.. J . Phys. Chem. 63, 226 (1959).
(9) Criegee. R . . Ann. 522, 75 (1936).
(10) Dintenfass, L., Kolloid-Z. 161, 60, 70 (1958).
(111 Dralev. J . E.. ArPonne ,Vatl. Lab. News-Bull. 3. 3 (1962).
(l2j Epstein. H . T.,
j . Phyr. ColloidChem. 54, 1053(1950).
(13) Fieser. I,. F., Fieser, .M., Organic Chemistry. p. 61,
Heath, Boston, 1950.
(14) Frenkel, ,J., Kinetic Theory of Liquids, pp. 366-426,
Dover Publications, New York, 1955.
(15) Gebhardt, J.: Herrington, K., J . Phys. Chem. 62,120 (1958).
(16) Gray! T. J., McCain. C. C.. Masse, N. G., Zbid., 63, 472
(19-lO~.
\-.--,.

McCormick, J. L., Baer. E.. J . ColloidSci. 18, 208 (1963).

Mathieson, R. T., Vature 183, 1803 (1959).
Moroney, IM. J.. Facts from Figures, pp. 96-107, Penguin
Books: Baltimore, 1962.
(22 Porter, K. R., Kallman, F., Exptl. Cell Res. 4, 127 (1953).
(231 R eyerson. L. H.: Honig, J. M.: J . Am. Chem. Soc. 75, 3917,
~

3020 (1057).

(24) R i & H.E., Jr., Kimball, W.A . , J . Phys. Chem. 59, 94 (1955);
Nature 181, 901 (1958).
(25) Sandler. Y . L.,J . Phvs. Chem. 58. 54. 58 (1954).
(26) Tammann, Y., ThLStates of Aggregatidn, Van Nostrand,
New York. 1925 (Engl. transl.).
P. A., 2. Elektrochem. 48, 675 (1942).
olmer, M., Ibid., 35, 555 (1929).
(29 Volmer, M.. Weber, A., Z . Phyrik. Chem. 119, 277 (1926).
(301 W ade, W. H., Hackerman, N.. J . Phys. Chem. 6 5 , 1681
(1961).
RECEIVED
for review July 22, 1963
ACCEPTED
November 5, 1963
30th Annual Chemical Engineering Symposium. ACS Division
of Industrial and Engineering Chemistry. University of Maryland.
November 1963. Based on Ph.D. thesis of G. D. Cheever.

(17) Gulbransen. E. A , : McMillan, W. R., Andrew, K. F.,

Trans. AZME 200, 1027 (1954).
(18) Harkins. \V. D.: Gam, D. &f.>J . Piiys. Chcm. 36, 86 (1932).

MECHANISM OF HEAT TRANSFER T O A

FIXED SURFACE IN A FLUIDIZED BED

EDWARD N .

Z I E G L E R A N D W I L L I A M T. B R A Z E L T Q N

The Technological Institute, IVorthwestern Crniiersity, Ecanston, Ill.

Simultaneous heat and mass transfer from the surface of a sphere was studied for comparable situations
in a gas stream and a gas fluidized bed of solid particles. The systems were chosen so that the fluidized
particles would have capacity for heat transport but not mass transport. Solid particles in the fluidized
state increased heat transfer 10- to 20-fold, but increased mass transfer only 1 /z to 2 times. On the basis
of the particles unique property for heat transport, and assuming analogous heat and mass transfer in the
gas phase, it was concluded that in any mechanism of heat transfer in the fluidized state, 80 to 95% of the
transfer must be accounted for by particle transfer mode. The remainder may be accounted for in Q path
solely in the gaseous phase.
HE phenomenon of heat transport in fluidized beds has
Tbeen
the subject of numerous studies in recent years.
Of particular interest is the heat transfer between the bed
and an internal or external surface. T h e results of work on
this topic have been summarized extensively by Zenz and
Othmer (26). Another review by Botterill (7) includes all
but the most receit literature.
Several of the early investigators verified the improvement
in heat transfer, and offcred accompanying rationales in the
form of mechanisms of transport. For instance, Leva
70) and Docv and Jakob (3) suggested
and coworkers (8--that the increase in heat transfer was probably a consequence of the scrubbing action of particles against the transfer surface. This action was thought to disturb the gas film
and hence decrease its resistance to the flow of heat. I n
later presentations. such as those of Van Heerden, Kobel,
and Van Kreveltm (78: 7 9 ) : \\*icke and Hedden (251, Wicke
and Fetting ( 2 4 ) ;Mickley and Fairbanks ( 7 7 , 717), and Ernst
( 4 ) . rnechanisms were developed that appear to be more
credible in explaining heat transfer improvement of the
magnitude experienced. Although there are some differences
in these siiggested models, common factors may be summarized
as follokzs. T h e fluidized particles are visualized as a packeti.e.. closely locked assemblage of particles-moving
from the
I

Presrnt addrrss. Esso Rpsearch and Engineering Co., Linden.

N. J .
94

l&EC

FUNDAMENTALS

core of the bed to the particular boundary surface, absorbing

or giving u p heat, depending upon the relative temperature
of the surface, and then returning to the core of the bed.
The interstitial gas serves as a stirring agent and as a heat
transfer medium between the particles and the surface.
T h e presence of the particles probably improves heat
transfer by a combination of the two aforementioned effects-.
conveyance of heat from the packet to the surface on contact
and disturbance of the gas film adjacent to the surface. The
simultaneous occurrence of these two effects has not yet been
demonstrated by experiment. I t was the intention of this
experimental investigation to verify the presence of these
two effects and to determine their relative importance.
T h e experiments performed involve simultaneous heat
and mass transfer from a solid object to a fluidized bed. T h e
system was chosen to allow both the particles and the gas to
take part in the transport of heat a t the solid surface. O n
the other hand, the particles in this system have negligible
absorptivity for the diffusing species and consequentlv have
no capacity for mass transport. Therefore! the only mechanism of importance in mass transfer is the diffusion of mass
through the surface gas film. In systems without fluidized
particles a t constant surface temperature and composition.
it is known that the mechanisms of heat and mass transfer
from a surface are analogous for low mass transfkr ra.tes.
Both types of transfer may be considered to take place through

L E A V I N G AIR
TO BE A N A L Y Z E D
B Y DEW POINT DEVICE
THERMOCOUPLE
READING OF
CELITE SURFACE
TEMPERATURE

I 1
LUCITE COLUMN
4 -DIAMETER

I
FLUIDIZED
P4RTlCLES

C E L I T E SPHERE
-DIAMETER

--

in the nearl) saturated air a t the interface may be assumed

equal to the vapor pressure a t the surface temperature. T h e
heat transfer coefficients are calculated from a knowledge of
the air temperature and surface temperature and the mass
transfer rate as determined by dew point measurements.
iZllowance is then made for radiant heat transfer. T h e
transfer area is that of the sphere minus the area of the thermocouples. T h e d a t a for the single sphere with no particles
present agree well with those of Ranz and Marshall (74)
and Griffith ( 7 ) .
Analysis

T h e heat and mass transfer under steady-state conditions

may be defined by the following equations:
For heat and mass transfer balance a t the interface

LYX

POROUS P L A T E

For mass transfer rate

ENTERING AIR AT
CONSTANT HUMIDITY

Figure 1, Experimental apparatus for simultaneous

heat and mass transfer from sphere to air with aid of
fluidized particles
a gas film near the solid surface. As a consequence of the
analogoris mechanisms for these conditions, the ratio of heat
to mass transfer coefficients remains approximately constant,
and there exists a n equality of j factors (2) for both cases.
L p o n the addition of fluidized particles to a system, the heat
transfer coefficient can be improved significantly.
The improvement of the mass transfer coefficient is studied
herein and is used in the determination of a mechanism of
heat transfer. For example, if heat is not transferred via the
particles. the mass transfer coefficient would no doubt increase
by the same factor as the heat transfer coefficient. O n the
other hand, if the particle does have a part to play in heat
transfer and not in the mass transfer, the transfer factors for
the tlvo cases ~vouldbe expected to be affected differently.

Experimental Techniques

T h e experimental equipment used is indicated in Figure 1.

A half inch-diameter sphere is saturated with distilled water
and then exposed to a metered stream of dry inlet air under
various fluidized bed conditions. T h e sphere was formed
from Johns-hlanville Iype VI11 catalyst carrier, a claylike
material which after baking has been shown experimentally
to have a convenient constant-rate drying period (77). T h e
solid sphere has a fine hole drilled through all but V 3 2 inch of
its thickness. A thermocouple is inserted in this hole and held
there by the addition of more clay and rebaking. T h e sphere
is then saturated with water and suspended from the thermocouple. As a consequence of the short height of suspension and
the relatively low gas velocities, the sphere remains stationary.
Temperatures are recorded throughout each run. T h e inlet
and outlet air hurnidities are recorded on a General Electric
dew point recorder (6) which has been calibrated to 1 F.
Temperatures of the air a t inlet and outlet conditions are
measured by thermocouples. When steady state has been
achieved, the outlet humidity and sphere temperature remain
constant until the falling rate period is reached. Copper,
glass? and Alundum particles were used as the fluidized
medium.
T h e mass transfer coefficients are calculated from a knowledge of inlet and outlet dew points and the surface temperature of the sphere. T h e partial pressure of the water vapor

For heat transfer rate

9 = h t ( t , - tJ7n

(3)

T h e quantities pp and t u vary little during the run. I n addition, p, is a function of t , only. T h e coefficients k , and h l
are independent variables and, depending 011 their values, a
certain value of t,, and hence p , , will be reached a t steady
state-i.e.. the final value of the temperature a t the surface is a
function of only the heat and mass transfer coefficients when the
dry bulb temperature and the partial pressure of the water
vapor in the air have small variation. T h e coefficients /io
and h , for most systems have been found to vary similarly.
but they should not be thought of as mutually dependent.
T h e transfer of heat and mass is interdependent? but the
coefficients of heat and mass transfer are independent. Only
under conditions a t which heat and mass are transported by
a n analogous mechanism d o the k, and h l have similar variation. For a n air-water system, Ranz and Marshall (74)
presented the following relation :
Nu

+ 0.60 Re1%c1/3

(4)

Equation 4 was developed from boundary layer theory as per

Froessling (5) and from dimensional analysis. It is applicable for mass transfer in the Reynolds number range
0 to 200. T h e analogous heat transfer equation with S u
and Sc replaced by Nu and Pr. respectively, is found to hold
equally well in the same range.
Results

Experiments were performed without fluidized particles

and the results were checked with Equation 4. T h e equations

Nu = 2

+ 0.56 R ~ * S C ~ ~

(Sa)

for mass transfer (1 1 .5y0standard deviation of data). and

N u = 2 f 0 68 Re1*Pr1/3

(jb)

for heat transfer (8.5% standard deviation) were used to represent the d a t a of this system over the range investigated.
With the addition of fluidized particles to the system.
however, a n entirely ne\\. mechanism for heat transfer. particle
transport (convenient name for transport via a path or paths
which include particles), is available which has no equivalent
in mass transfer for the conditions of the experiment ThereVOL. 3

NO. 2

M A Y

1964

95

fore, for the system with fluidized particles, it is conceivable,

and in fact a reality, that the heat transfer coefficient can
change significantly from that of the no particle case with
no comparable change in mass transfer coefficient. Table
I is a sampling of representative heat and mass transfer coefficients and their ratios for the fluidized systems and empty
tubes. For all systems of fluidized particles. a large increase
in heat transfer coefficient occurs with a comparably minor
increase in mass transfer coefficient. The ratio h,/ku is approximately constant (0.249, 4% standard deviation) for
nonfluidized systems. This is in agreement with the Lewis
relation (after W . K. Lewis) which states that h,/k, is approximately equal to C, and is a consequence of the analogous
mechanism of heat and mass transport. It has been shown
that for the air-water system, the relation is applicable. For
pure air, C, = 0.24.
The ratio h,/kv is plotted in Figure 2 as a function of gas
velocity over the entire experimental range. T h e variation
in the ratio, thus represented, is slight. With fluidized particles present, however, the ratio varies considerably and is
greater than 5.0 in some instances. I t is apparent that the
mechanism of heat transfer is not analogous to that for mass
transfer in fluidized beds. Also, Equation 5 does not apply
for systems with fluidized particles. The values of heat and
mass transfer coefficients are plotted as a function of reduced
gas velocity in Figure 3.

It is of interest to determine the effect, if any, of the moving

particles on the gas film mass transfer. T o compare coefficients in systems before and after the addition of particles.
it is essential that the comparison be made at identical average
gas velocities so that changes are not caused merely by a
change in velocity. Therefore, the fluidized bed coefficient,
k,, is compared with ku, defined as the coefficient before fluidized particles are added but a t the same average gas velocity,
G,, as is present in the fluidized bed. The coefficient ku
may be determined from Equation 5, which represents the
no particle case for this experimental work using G, in the
Reynolds number. In each case k , is greater than k,, The
factor of improvement, kv/kv, varies little in all systems studied,
the highest value being about 2. Representative values are
shown for each of the systems in Table 11.
The heat transferred to the interface may be considered as

0 23
022

I20

130

I40

150

160

180

I70

190

200

G or GI

Figure 2.

Ratio h./ky vs. gas velocity

210

220

(hf/h,)

(8)

In the experiments with no fluidized particles present it was

previously shown that the ratio h,/ku has little variation with
change in gas velocity past the surface (see Figure 2). This
is in accord with the heat and mass transfer analogy for the
air-water system. For systems containing fluidized particles
the ratio of hf/k, is assumed to vary similarly. since it represents
the ratio of coefficients for the gas phase where the analogy
continues to exist. In fact, this ratio should be the same as
h,/ky in the no particle case evaluated at the same average
gas velocity. Hence, hrniay be calculated from:
h,

X A-ALUNDUM
B 8-ALUNDUM

(7)

Neglecting radiation from the fluidized system surrounding

the sphere, a reasonable assumption considering the temperatures of the experiments, h c becomes equal to h,. By combining
Equations 3, 6, and 7, the fraction of the total heat transport
which is attributed to particle transport may be represented as:
4P/Y = 1

020
0

(6)

YI = hf(t, - d m

0 2 7 c

+ Yr

where yp is the heat transferred via particle transport and

q f is the heat transferred via the continuous gas phase. In
terms of its coefficient

V.L.2

xz

= YP

(h,,k,)k,

(9)

where h,/kl is obtained from Figure 2, using the average gas

velocity in the fluidized bed, G,, for the gas velocity. In-

0 COPPER

A GLASS

I .o

X A-ALUNDUM
8 8-ALUNDUM

0 COPPER
A GLASS

I-

5u -

IO

iuL .=
540W

sS30-

.>
2- IO
v)

---e

/A&--&

&*
:Ezo-

A
4x

071
0.6

Table I.

D p f X 703,

Fluidized
Particles

In.
...

None

2.3
2.3

Copper powder
A!lundum (sharp)

5.8
5.8
3.5
4.5

13
Glass spheres

Representative Experimental Data

Lb.
G Hr.-Sq. Ft.

h,,

G/GW

145.6
163.2
70.0
145.6

3.18
6.60

92.6
145.6
92.5
72.3
124.7

2.27
3.57
5.78
3.61
6.23

...

Lb.
ku Hr.-Sq. Ft.
H I . -OF.-Sq. Ft.
Mole Fraction
3.12
12.3
3.79
15.6
34.0
21.4
34.0
26.4

B.t.u.

3.48
2.99
2.63
4.22
5.40

21.9
24.5
23.7
23.9
18.6

76.0
73.3
62.4
100.9
100.5

Relative Effect of Particles on Film (Representative Values)

Lb.
Lb.
B.t.u./Hr.-Sq. Ft.-F.
Hr.-Sq. Ft. Mol. Fract.
Ge Hr.-Sq. Ft.
h,
hc
h/
kU
k,
131.7
34.0
3.25
5.29
21.4
13.1
172.2
42.8
3.62
5.76
23.0
14.5

hclku
0.253
0.244
1.59
1.29

Table II.
Flutdtred
Pmttcles

Copper
Alundum
A
..

13

Void
Fraction,

0.532
0.632
0 784

n 706
0 722
0 751

Glass

0.570
0.618

159 . O
182 9
12s 0
144 6
126.9
150.0

73.3
73 3
62 4
57 9
100.9
81.1

3.51

6.15
6 11
5 90
5 30
5.92
4.51

3 71

3 22
3 37
3.22
3.43

terpolation between the discrete experimental values of gas

velocity is necessary in using Figure 2, but an exact value of
velocity is not necessary because very little change occurs in
the ratio hc/kv. Using Equation 9 to compute h l and the
experimental value of h,. the ratio q p / q may now be calculated
from Equation 8. Figure 4 is a plot of this ratio us. reduced
mass velocity of the gas. This indicates that 80 to 95% of
heat transfer, neglecting radiation, is by the particle transport
mechanism ; the actual percentage appears to depend on
particle properties. Assuming that Equation 9 is correct,
i t is apparent that the factor of improvement of heat transfer
in the gas film, h1,hc, is exactly the same as the improvement
factors for mass transfer, ky,ky, listed in Table 11. Also included in Table I1 is the per cent of the total heat transfer
caused by the film disturbance ( h l - h,)/h, X 1000/0 for these
systems. These values are less than 8y0in all systems tested.
The heat transfer coefficient is dependent upon the kind of
fluidized particles used, whereas mass transfer coefficients
seem to have little dependence of this type in the range studied.
If heat )transfer by particle transport is accepted as the dominant
mechanism, it is understandable that a particle property such
as spec:ific heat and a geometry factor such as particle diameter might be of considerable influence. In a number of
articles ( 7 7 , 78, 200-2.3) it was concluded that h, increases
with increasing density and specific heat of the solid particles.
T h e exact relationship between h, and the solid properties
varied among the different authors. T h e probable cause
of this variation was the different conditions of gas flow rate
and solid materials investigated. Likewise, an explicit
depency of h , on particle diameter, valid for all conditions,
could not be found because of the differences in experimental
range and the effect of particle diameter on bed porosity.
In the work of Sarkits, Traber. and Mukhlenov (73> 75, 7 6 ) )
a correlation of experimental data is divided into two regions
of f l o i v ithe laminar flow regime and the turbulent flow regime). T h e direct dependency of h, on solids density and

24.7
24 5
23 7
27 0
23.9
18.1

14.1
14.9
13 0
13 5
13.0
13.8

1.6
1.6

6.0
5.0

1.8
1.6
1.8
2.0
1.8
1.3

3.6
3.3
4.3
5.8
2.7
1.3

specific heat becomes weaker as the flow becomes more turbulent. Nevertheless, h, increases with increasing values of
these properties in both regions. The particle diameter
effect differs for the two regions, however. For laminar
flow, h, varies inversely with particle diameter, and for turbulent flow, it has direct variation. Also, higher coefficients,
h,, resulted for smooth particles than for coarse particles.
A more detailed description of the conditions necessary for
laminar and turbulent flow is given by Sarkits, Traber, and
Mukhlenov (76).
In the present investigation, the smooth glass spheres of
highest specific heat gave the highest values of h, and the
sharper copper powder of lowest specific heat, the lowest h,
values. Two sizes of Alundum particles were used and gave
little apparent variation in h,. T h e values of h, vary with the
solid properties in a manner consistent with the previous
articles cited. The depth of suspension of the sphere may be
another variable of importance.
For the region of gas velocities and particle properties studied, the predominant mechanism of heat transfer to a surface
in a fluidized bed is that of particle transport. The other
mechanism, additional disturbance of the gas film, improved
heat and mass transfer, but the heat transferred in this manner
is only a small portion of the total.
Nomenclature

CYl

D,
DPJ
Do
G
G,
h,

= heat capacity of gas, B.t.u./lb.- F.

= diameter of sphere, ft.
= diameter of fluidized particles, in.
=

gas diffusivity, sq. ft./hr.

= superficial gas velocity, lb./hr.-sq. ft.

= average gas velocity based on void area of bed:

lb. hr.-sq. ft.

radiation heat transfer coefficient, B.t.u./hr.sq. f t . - O F.
= total heat transfer coefficient, B.t.u.,/hr.-sq. ft.=

F.
VOL. 3

NO. 2

MAY

1964

97

- h , = experimental heat transfer coefficient

of mechanisms other than radiation. B.t.u. hr,sq. f t . - O F.
= heat transfer coefficient for film convection in
presence of fluidized particles, B.t.u./hr.sa. ft.- F.
= heat transfer coefficient for film convection in
absence of fluidized particles, evaluated a t G,
of fluidized bed, B.t.u./hr.-sq. ft.-O F.
= mass transfer coefficient, 1b.-moles/hr.-sq. ft.atm.
= experimental mass transfer coefficient, kuPaMu=
lb./hr.-sq. ft.
= mass transfer coefficient in absence of fluidized
particles, evaluated from Equation 5 a t G,
of fluidized bed, Ib./hr.-sq. ft.
= molecular weight of air, lb./lb.-mole
= h,Dp/o
= k,D,/pD,
= av. partial pressure of air, atm.
= saturation partial pressure of water vapor,
vapor pressure corresponding to t,, atm.
= partial pressure of water vapor in the air, atm.
= logarithmic mean partial pressure difference,
atm.
= ht

hc
h!

= Cpp,cy
= rate of heat transfer a t interface, by particle

transport, via gas film, B. t.u.:hr.-sq. ft.

GD,/p, Reynolds No. based on superficial gas
mass velocity
= dpDr
= bulk gas temperature, temperature a t surface of
sph;ere, O F.
= loearithmic
mean temuerature difference. O F

= mass transfer rate, 1b.-moles,hr.-sq. ft.

=

GREEK
cy

= gas thermal conductivity, B.t.u./hr.-ft.-

bed void fraction, dimensionless

= gas viscosity, lb./hr.-ft.
= gas density, Ib./cu. ft.

(1) Botterill, J. S. M., Brit. Chem. En!. 8, 21 (1963).

(2) Chilton, T. H., Colburn, A. P., Ind. Eng. Chem. 26, 1183
(1934).
(3) DO, CV. M., Jakob, M., Chem. Eng. Progr. 47, 637 (1951).
(4) Emst, R., Chem.-Ingr.-Tech. 31, 166 (1959).
(5) Froessling, N., Geriunds Beitr. Geophyr. 52, 170 (1938).
(6) General Electric Co., Dew Point Recorder Pamphlet,
GEI-40444A.
(7) Griffith, R. M . , Chem. Eng. Sci. 12, 198 (1960).
(8) Leva, M., General Discussion of Heat Transfer, London,
SeDtrmber
1951.
.
.~
(9) Leva, M., Grummer, M., Ind. Eng. Chem. 40, 415 (1948).
(10) Leva, M., Lleintraub, M., Grummer, .M,, Chern. Eng. Ptogr.
4 5 , 563 (1949).
i l l ) Micklev. H. S..Fairbanks. D. F.. i4.I.Ch.E. J . 1. 374 1195.5).
(12) Mickley; H. S., Fairbanks, D. F., Hawthorn, k.D.: Chem
Ene. P?OPY.
.Cvmb. S c r . 32. 51 11961\.
(13) uMukhenov, I. P.,- Traber, D. G., Sarkits, V. S.,Bondarchuck, T. P., Zh. Priklud. Khim.32, No. 6, 1291 (1959).
(14) Ranz, W.E., Marshall, LV. R., Jr., Chem. Eng. Progr. 4 8 , 141,
173 (1952).
(15) Sarkits; V. B., Traber, D. G., Mukhlenov, I. P., Zh. Ptiklad.
Khim. 32, No. 10, 2218 (1959).
(16) Z6id.,33, N o . 10, 2197, 2200 (1960).
(17) Thodos, G., Northwestern University, Evanston, Ill.,
personal communication.
(18) Van Heerden. C., Nobel, A. P. P., Van Krevelen, D. LV.,
Chem. Ene. Sci. 1. N o . 2. 51 11951).
(19) Van keerdeh, C., hobel, A: P. P., Van Krevelen, D. LV.,
Ind. Eng. Chem. 4 5 , 1237 (1953).
(20) Vreedenberg, H. A , , Chem. Eng. Sci.11, 274 (1960).
(21) Vreedenberg, H. A,, J. .4fipl. Chem. 2, Suppl. Issue No. 1.
S26 11952).
(22) Cqen, C.-Y.,Leva, Max, A.I.Ch.E. J . 2, 482 (1956).
(23) LlendFr, L., Cooper, G. T., Ibid.,4 , 15 (1958).
(24) bicke, E., Fettins, F., Chem.-In,nr.-Tech. 26, 301 (1954).
(25) bVicke, E., Hedden, K.: Ibid.,2 4 , 82 (1952).
(26) Zcnz, F. A . , Othmer, D. F., Fluidization and Fluid Particle Systems, Reinhold, New York, 1960.
~~

\ - -

F.

iL

literature Cited

B.t.u.
latent heat of vaporization of water, _ _
lb.-mole

RECEIVED
for review July 25, 1963
ACCEPTED
November 22, 1963
Lliork performed under the auspices of the I!. S. Atomic Energy
Commisqion at the Argonne National Labordtorv, Argonnc, Ill.

CONCENTRATION AND MASS FLOW

DISTRIBUTIONS IN A GAS-SOLID
SUSPENSION
S . L . S O O , G . J . T R E Z E K , R . C . D I M I C K , A N D G. F . H O H N S T R E I T E R
University of Illinois, Crbana, Ill.

Distributions of concentration, mass flow, and velocity of solid particles were studied with a fiber-optic probe
and an electrostatic probe. Concepts concerning these distributions and electrostatic charges on solid particles were furthered and substantiated. The relation between electrostatic charge on solid particles and
diffusivity of solid particles, and the difference between static loading and mass flow ratio of phases were
proved.

and correlation of momentum, heat, and mass

transfer in a gas-solid suspension call for a knowledge of
the concentration distribution of solid particles and the velocities of the phases. Earlier studies are applicable to given
concentration distributions ( 7 7 . 79) or assume constant
concentration in analysis (22) and data correlation from overall average concentration (3, 27). Measurements were made
by impact counter system with the assumption of similar
REDICTION

Present address, Lockheed Research Laboratory, Palo Alto,

Calif.
98

I&EC

FUNDAMENTALS

velocity of phases ( 7 8 ) ; the differences between velocity of

phases and between mass flow and concentration distribution of
particles were later recognized and presented (73). These,
together with studies on dense suspensions (15), measurement
by point study probe ( 2 3 ) , and measurement by capacitance
(24), show that more definitive measurements need to be
made. Results of earlier studies (73, 78) are further generalized in the study of boundary layer theory of gas-solid
suspensions ( 7 6 ) .
Methods of measurement of concentration of the particulate
phase in a suspension consist of electric measurement as