Sdarticle
Sdarticle
Sdarticle
David G. Thomas
Oak Ridge National Laboratory, Oak Ridge, Tennessee
Received October 14; 196~, revised December 6, 196~
ABSTRACT
A critical analysis was made of the extensive experimental data on the relative
viscosity of suspensions of uniform spherical particles. By appropriate extrapolation
techniques, non-Newtonian, inertial, and nonhomogeneous suspension effects were
minimized. As a result, the scatter of the data was reduced from ± 7 5 % to ± 1 3 % at a
volume fraction solids of 0.50. The coefficients of different power series relating rela-
tive viscosity and volume fraction solids were determined using a nonlinear least
squares procedure. It was shown that a new expression containing three terms of a
power series with coefficients determined from previous theoretical analyses and an
exponential term with two adjustable constants fit the data as well as a power series
with six terms, either three or four of which were adjustable constants with the re-
maining coefficients being theoretical values.
INTRODUCTION
Previous papers in this series have been largely concerned with the
transport characteristics of non-Newtonian suspensions (1). One of the
major problems encountered in studies of such suspensions in the deter-
ruination of the limiting viscosity at high rates of shear (2). This Value is
necessary for use in turbulent transport correlations (1) and for verifica-
tion of asymptotic values predicted by different rheological equations of
state (2). In either case the viscosity of dispersed Newtonian suspensions
is of special interest, because it represents a mininmln value for the limiting
viscosity of non-Newtoniaa suspensions.
Since Einstein published an analysis for the viscosity of dilute suspen-
sions (3) in 1905, one of the most challenging rheological problems has been
the development of theoretical or empirical expressions for the viscosity of
concentrated suspensions. The results of these studies have been sum-
marized in recent review articles (4-6). As yet, there is no commonly ac-
cepted theoretical treatment for concentrated suspensions; perhaps the
1 Research jointly sponsored by The Office of Saline Water, U. S. Department of
the Interior, and U. S. Atomic Energy Commission under contract with the Union
Carbide Corporation.
267
268 THOMAS
principal reason for this is the difficulty of the analysis coupled with the
poor agreement among different experimental investigations (even though
the majority of these studies were carried out under carefully controlled
conditions). The object of this note is (I) to propose criteria for isolating
hydrongmic interaction of particles independent of such factors as floccula-
tion, particle size, particle size distribution, and type of viscometer and
(2) to apply these criteria to the experimental data and the evaluation of a
consistent set of parameters for several different expressions for the sus-
pension viscosity.
THEORY
Extreme Dilution
The theoretical calculation of the viscosity of suspensions of solid par-
ticles can be approached in several ways. First, the viscosity may be deter-
mined either from the velocity gradient (4, 7)
{OU, OU~
r,k = --~ \ ~ + b--~J' [1]
the fact that for spheres in contact the relative viscosity must be infinite,
Mooney suggested a value 1.35 ~ a2 ~ 1.91.
Ford (16) advanced the proposition that the experimental data may be
fitted more accurately by an expression of the form:
~0/~ = 1 -- 2.,~b -t- b~ 5 -- b~4)7, [6]
where values of the constants were taken to be b2 -- 11.0 and b3 = 11.5
oil the basis of Vand's data. Ford also suggested that the 4)5 term becomes
important at the onset of the inhibition of particle rotation and the 4)7
term becomes important at the onset of particle interlocking.
The finite size of the particles which effectively shields the central particle
from interaction with other than nearest neighbors at the concentration is
increased has been treated by Simha (15). The effect of the shielding re-
duces the result calculated on the assumption that the particles are points.
To obtain a solution to the hydrodynamic equations, the cage of particles
is replaced by a concentric spherical enclosure. As a result, the model is
only approximate for dilute concentrations. For dilute concentrations, the
result can be expanded to
~/#0 = 1 + 2.5¢ [1 + 25¢/4 f3 . . . ] , [7a]
whereas for concentrated suspensions the relative viscosity is approximated
by
lira ~/~0 = 1 -{- ( 5 4 / 5 f ) [ 4 ) 2 / { 1 - (¢/4) .... )}3], [75]
~b-*¢ m a x
where 1 < f < 2 but may be expected to vary within those limits over the
complete range of concentrations. It should be emphasized that Eqs. [7a]
and [75] are approximate expressions the use of which should be limited to
conditions of moderate ( 4 ) ~ 0.10) and very large concentrations, respec-
tively (35).
DATA ANALYSIS
Data from many different investigations of the viscosity of suspensions
of spheres are shown in Fig. 1. The data show a spread of from about :i:20 %
at ~ = 0.20 to about :i:75 % at 4) = 0.5. These data were obtained with
both rotational and capillary viscometers and represent, a range of particle
diameters from 0.099 to 435 microns. In the majority of cases, the particles
were rather closely sized. Since the data given by Eveson (19) show that
mixtures of spheres having a fourfold variation in diameter have less than
a 6 % difference in relative viscosity up to ~ = 0.20, it is believed that the
variation about the mean particle diameter in the investigations cited above
contributed very little to the scatter of the data. The particle materials
included polystyrene, rubber latex, glass, and methyl methacrylate. In all
TRANSPORT CHARACTERISTICS OF SUSPENSIONS 271
t0 2
REFERENCE
e I
O 25
[] 26 I
A 27
V t5
/
• 12 /
• 24 j B~
& 28
• 20 /' I
29 /
OD 25
(D 21
Ill 19 c~
2 in 30 /
=t pl 22 ~ 0 /
>_- G
'¢'
5!
3z /
B ()
o /
/ /,
10 Be /
I--
/
/
/ pr "'
/
in
I ~' o /
'• in
,,,fi ,~# ,o /
>7
f Fo-
• u/3
. o ~ ~'°
0 0.1 0.2 03 0.4 0.5 0.6 07 0.8
VOLUME FRACTION SOLIDS,
studies, either the density of the suspending medium was adjusted or the
viscosity of the suspending medium was sufficiently large that settling was
unimportant. Examination of the experimental procedure used in these
studies shows no basis for eliminating any of the data because of faulty
technique; consequently, there nmst be at least, one additional parameter
that has not been accounted for.
272 THOMAS
10 2
EQ, 70 AND 7b
WlTt TABLE I
. . . . EQ. 9
>.-
I-.
0
:
40
> #
laJ 8,, /0
>
I-
<Z
_1 6
laJ
cl:
./
~s
jv
,J'
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
VOLUME FRACTION SOLIDS,
Bruijn (17) believes that inertial effects due to the restoration of particle
rotation after collision result in an additional energy dissipation and a
consequent increase in relative viscosity with increasing particle diameter.
In flow through capillary tubes, the increase in viscosity observed with
large particle size suspensions is opposed by a decrease in viscosity caused
by a tendency for particles to migrate toward the center of the tube as the
particle diameter is increased (18).
Examination of the data from which Fig. 1 was prepared showed that
in several cases the tests covered a sufficient range of shear rates or particle
sizes that it was possible to extrapolate to conditions where particle size
effects were negligible. For particles less than 1 micron diameter, the limit-
ing value of the relative viscosity was obtained as the intercept of either a
linear plot of 1/Dp versus u~/uo or a linear plot of 1/(du/dr) versus #.~/uo.
For particles larger than 1 to 10 microns, the limiting value of the relative
viscosity was obtained as the intercept of a linear plot of Dp versus u,~/uo •
In the event that large particle size data were also available as a function
of shear rate, the reduced particle size data were further corrected by
plotting against 1/(du/dr). T r e a t m e n t of the suitable data in this manner
gave a unique curve for which the m a x i m u m deviation was reduced from
three- to sixfold over t h a t shown in Fig. l, that is, to 4 - 7 % at 4, = 0.20
and to 4-13% at ~ = 0.5 (Fig. 2).
COEFFICIENT EVALUATION
TABLE I
Coefficients for Eq. [7b]
(P p~/Po f
TABLE II
Coeffcients of Power Series
Iz8
IZo
Case ~ A2 A3 A4 As A6 A7 Variance
r a n d o m l y p a c k e d b e d of u n i f o r m s p h e r e s d e t e r m i n e d b y e x t r a p o l a t i o n t o a
v a l u e of z e r o f o r t h e r a t i o of s p h e r e d i a m e t e r t o c o n t a i n e r d i a m e t e r ( 3 3 ) .
W i t h (~max ~ - 0 . 6 2 5 , t h e v a l u e s of f i n E q . [7b] w e r e d e t e r m i n e d ; s e l e c t e d
values are given in Table If
As pointed out b y Simha a n d Somcynsky (35), Eqs. [7al and [7b] were applied
over a greater c o n c e n t r a t i o n range t h a n i n t e n d e d originally (15). A more exact appli-
cation is t r e a t e d elsewhere (35).
TRANSPORT CHARACTERISTICS OF SUSPENSIONS 275
DISCUSSION
sizes and shear rates with the object of minimizing secondary effects such
as those due to non-Newtonian behavior, inertial forces, and measuring
instrument wall effects. None of the data from Fig. 1 which were suitable
for extrapolation were eliminated in the preparation of Fig. 2. On the other
hand, because of the requirement for extrapolation to obtain a unique set of
data in which secondary effects were minimized, the results of several very
thorough investigations could not be included. This emphasizes the im-
portance of careful planning of future experiments to insure that a suffi-
cient range of conditions is covered to permit evaluation of the secondary
effects.
ACKNOWLEDGMENT
The author wishes to acknowledge the assistance of F. L. Miller, Jr., who per-
formed the nonlinear least squares analysis.
NOTATION
E~ = viscous dissipation per unit volume, ergs/crn. 3
f = shielding factor, dimensionless.
U = velocity, cm./sec.
V = volume, elYI . a
Subscripts
, = suspension.
0 = suspending medium.
niax = m a x i m u m value.
Greek Letters
= volume fraction solids, dimensionless.
r = stress, dynes/cm. 2
= viscosity, dyne-sec./cm. 2
~/~ = mean value of relative viscosity, dimensionless.
REFERENCES
1. THOMAS, D. G., A . I . C h . E . J . 6, 631 (1960); ibid. 7, 423,431 (1961); ibid. 8, 266,
373 (1962).
2. THOMAS,D. G., I n J. F. Masi and D. H. Tsai, eds., "Progress in International
Research on Thermodynamic and Transport Properties," p. 699. Academic
Press, New York, 1962.
3. EINSTEIN, A., Ann. P h y s i k 17, 459 (1905); ibid. 19, 289, 271 (1906); ibid. 34, 591
(1911).
4. FRISCn, H. L., .~NDSIMHA,R., I n F. R. Eirich, ed., "Rheology," Vol. 1, Academic
Press, New York, 1956.
5. HERMANS, J. J., Ed., "Flow Properties of Disperse Systems." Interseience, New
York, 1953.
6. RUTGERS,I. R., Rheol. Acta 2,202, 305 (1962).
TRANSPORT CHARACTERISTICS OF SUSPENSIONS 277