Agitation and Mixing

11. AGITATION AND MIXING

11.1 Introduction
Many processing operations depend for their success on the effective agitation and
mixing of fluids. Agitation differs from mixing in that the former usually refers to the
circulatory motion of a fluid inside a vessel, while the latter is associated with the random
distribution one into another of at least two phases (one of which may be a solid). The
homogeneity of the mix differs from one case to another. For example, mixing of gases
always produces a homogeneous phase where the composition is constant throughout the
mixture. On the other hand, mixing of cement and sand with water will not produce such
homogeneous mix.

11.2 Agitation of liquids

11.2.1 Purpose of agitation
Liquids are agitated for a number of reasons:
 Suspending solid particles.
 Blending miscible liquids (such as ethanol and water).
 Dispersing a gas in a liquid in bubble form.
 Dispersing a liquid into an immiscible liquid to form an emulsion.
 Increasing the heat transfer coefficient between a liquid and a cooling coil or
jacket.
Sometimes more than one function can be achieved: In hydrogenation of a liquid,
agitation serves to disperse hydrogen gas into the liquid besides helping to remove the
heat of reaction through the external cooling coil or jacket.

11.2.2 Agitation equipment

A typical equipment for agitation is shown in
Figure (11.1). It consists of a cylindrical vessel
fitted with a central shaft ending with an
impeller. The motor is either directly
connected to the shaft or through a belt drive.
The edges of the tank bottom are usually
rounded to eliminate regions where fluid
currents would not penetrate.
The tank may be fitted with an external jacket
for cooling or heating. Baffles are often fixed
to the vessel walls to prevent the formation of
vortices.

Fig (11.1) Agitated vessel

128
 Agitation and Mixing

11.2.3 Impellers
In general, there are three main types of impellers for liquid agitation purposes.
Propellers
A propeller is an axial flow, high speed impeller for liquids of low viscosity. Their speeds
range from 400 rpm in case of large propellers to 1700 rpm for smaller sizes. The flow
currents leaving the impeller continue through the liquid in a given direction until
deflected by the floor or walls. The propeller blades vigorously shear the liquid causing
high turbulence to prevail. Such impellers are particularly effective in large vessels.
Figure (11.2) shows a three blade marine propeller. The size of such propellers does not
usually exceed 18" regardless of the diameter of the vessel.

Fig (11.2): Three blade Propeller

Paddle agitators
These simply consist of a number of paddles (usually 2 or 4) mounted on a vertical shaft.
They are used to slow agitate moderately viscous liquids. The pattern of currents is
mainly radial with no vertical motion. Sometimes more than one set of paddles can be
fitted to the central shaft. These impellers usually rotate at relatively low speeds
averaging 80 rpm. Their size ranges from 60 to 80% of the vessel diameter while their
width is about 10% of their length. It is usual to use unbaffled vessels for low speeds (up
to 50 rpm), while for higher speeds vortices tend to from so that baffling is necessary.
Sometimes the shape of the paddles may accommodate that of the bottom of the vessel to
prevent scale deposition on heat transfer surfaces. One typical such type is the anchor
agitator which is used in agitating viscous liquids while providing excellent heat transfer
to an external cooling jacket. (Figure 11.3)

Fig (11.3): Anchor agitator

129
 Agitation and Mixing

Turbine agitators
These resemble multibladed paddle agitators except that they can turn at higher speeds.
They operate over a wide range of liquid viscosities. The blades may be straight, curved
or pitched (Figure 11.4). The impeller diameter ranges from 30 to 50% of that of the
vessel. In low viscosity liquids, the rapid currents formed possess both radial and
tangential components. These latter components cause the formation of vortices which
have to be broken since they do not result in proper mixing. On the other hand, the zone
near the impeller is a zone of high turbulence and intensive shear.

(a) (b) (c) (d)

Fig (11.4): (a) Open straight blade turbine impeller

(b) Bladed disk turbine impeller
(c) Curved blade turbine impeller
(d) Pitched blade turbine impeller

Figure (11.5) shows details of curved and pitched blades turbine impellers.

(a) (b)

Fig (11.5): (a) Curved blades and (b) Pitched blades turbines

130
 Agitation and Mixing

11.2.4 Pattern of flow

The type of flow depends on the type of impeller, fluid characteristics, size and geometry
of vessel and agitator.
The velocity of the fluid at any point has three components: A radial and a tangential
component in a plane perpendicular on that of the shaft and a longitudinal component
parallel to the direction of the shaft. The tangential component is responsible for the
formation of a vortex with little or no mixing while the two other components are
responsible for the mixing action.
In small vessels, this swirling action can be
prevented by mounting the shaft off the center
line of the vessel (Figure 11.6). In large vessels,
on the other hand, it is customary to use up to
four baffles welded to the vessel walls. The flow
pattern in these cases change radically since the
swirling vortices are not formed and the
efficiency of agitation is highly enhanced.
Fig (11.6): Eliminating swirling

11.2.5 Recommended dimensions for turbine mixers

Although the design of an agitated vessel vary considerably depending on the location of
the impeller, the number of baffles (if any), the relative proportion of the vessel, etc…it is
customary in case of turbine impellers to use some typical relative dimension. Figure
(11.7)

H
W

De E

Dt

De 1 E 1 L 1
S1   S2   S3  
Dt 3 Dt 3 De 4

W 1 J 1 H
S4   S5   S6  1
De 5 Dt 12 Dt

Fig (11.7): Typical turbine proportions

131
 Agitation and Mixing

11.3 Power consumption in agitation

11.11.1 Elaboration of an equation to predict power consumption
The power consumed in liquid agitation depends on several factors: speed of rotation (rps
= n), impeller diameter (De), fluid properties: density (ρ), viscosity (μ) besides vessel
geometry. This last factor is quantified by the relative ratios shown in Figure (11.7). A
general relation can be written in the form:
P  f ( n, De ,  ,  , g , S1 , S 2 , S 3 ,...) (11.1)
Where g = gravitational acceleration and S1, S2, ,… = dimensionless shape factors defined
in Figure (11.7).
The elaboration of a suitable formula for the prediction of power consumption in a
agitation operation, we use dimensional analysis as follows:
Let P  k . a . b .n c .g d .Deh (11.2)
Now, denoting by Θ, L and M the basic dimensions of time, length and mass respectively,
we get:
a b c d
M .L2 M   M   1   L  h
  3  .  .  . 2  .L
  L   L.       
3

Hence:
M .L2
  bc 2 d .L3a bd h .M a b
 3

Equating the powers of similar terms we get the following set of equations:
b + c + 2d = 3 (i)
– 3a – b + d +h = 2 (ii)
a+b=1 (iii)
Solving for a, c and h, considering b and d to be independent parameters, we get:
a=1–b (iv)
c = 3 – b – 2d (v)
h = 5 – 2b – d (vi)
Replacing in equation (11.2):
P  k . 1b . b .n 3b2d .g d .De52bd
Grouping alike powers, we get:
b d
P   .n.De2   n 2 .De 
 k .  .  (11.3)
 .n 3 .De5     g 

132
 Agitation and Mixing

  .n.De2 
The first dimensionless group  . in the RHS is a modified Reynolds number Re
  
where the usual term of velocity (v) has been replaced by (n.De), which is a variation of
the ω.R form.
 n 2 .De 
The second term   represents the ratio between the centrifugal acceleration n2.De
 g 
2
(to replace ω .R) and the gravitational acceleration (g). It is called the Froude number
Fr.
 P 
The LHS  
5  is a dimensionless group known as the Power number Po
  .n 3
.De 

It is analogous to the friction factor in fluid flow.

Equation (11.1) can then be put in the form:
Po = f (Re, Fr, S1, S2, S3, …) (11.4)

11.11.2 Power correlations for specific impellers

The various shape factors in equation (11.4) depend on the type and arrangement of the
equipment. The shape factors S1, S2, S3, S4, S5 and S6 are defined as follows with reference
to Figure (11.7):
De E L W J H
S1  , S2  , S3  , S4  , S5  , S6  (11.5)
Dt Dt De De Dt Dt
In addition, the number of baffles and the number of impeller blades must be specified.
Figure (11.8) shows a typical plot of Po against Re for baffled tanks fitted with centrally
located flat bladed turbines with 6 blades. Different values of shape factors are shown on
the figure for different curves.
We note that in all three baffled curves, the Froude number does not play any role in
assessing the Po – Re relation. This is since the presence of baffles prevents the swirling
motion which is associated with the appearance of Froude number. Curve D on the other
hand, denoted the case of unbaffled tanks. The portion following Re = 103 is dotted. This
is since the values of power number calculated thereof have to be corrected to take into
consideration vortex formation owing to high speed of revolution. However the use of
unbaffled tanks is very rare at high Reynolds numbers. That is why; correlations
involving the use of Froude number are not of interest.
It is to also be noted that the difference between curves A and C is that C represents the
case of pitched blade turbines while curve A corresponds to flat blades.

133
 Agitation and Mixing

Fig (11.8): Power number - Reynolds number plots for 6 bladed turbines

We also note that at low values of Re < 10, all curves have a slope = – 1 on the
logarithmic scale corresponding to a relation in the form:
KL
Po  (11.6)
Re
Hence:
P  .K L

 .n .De  .n.De2
3 5

Leading to the following equation for power prediction:

P  K L .n 2 .De3 . (11.7)
On the other hand, at high values of Re > 104, the values of Re tend to stabilize so that Re
= KT. Hence, the power can be calculated from the following equation:
P  K T .n 3 .De5 . (11.8)
The values of the constants KL and KT are tabulated in Table (11.1) for different types of
impellers for baffled tanks (4 baffles) with S4 ≈ 0.2.

134
 Agitation and Mixing

Table 11.1: Values of KL and KT

Type of impeller KL KT
Propeller three blades 41 0.32
Turbine
 Six blade disk 65 5.75
 Four pitched blade 411.5 2.27
Flat paddle (2 paddles) 311.5 2.7
Anchor 300 0.35

If the Reynolds number is in the range 10 – 104, then the previous equations (11.7) and
(11.8) can no more be of use and one has to predict the power consumption through
charts such as that shown in Figure (11.8).

Example 11.1
A six blades disk turbine impeller is installed centrally in a vertical baffled tank 2 m in
diameter. The turbine has a diameter = 670 mm and is positioned 670 mm above the
bottom of the tank. The turbine blades are 134 mm wide. The tank is filled to a depth of 2
m with an aqueous solution of 50% NaOH at 65oC (Density = 1500 kg.m-3, viscosity = 12
cP). Calculate the power required for a speed of revolution of impeller = 90 rpm. Then
deduce the power of the motor assuming an efficiency = 0.85.
Solution:
We have: n = 2.5 rps, De = 0.67 m, ρ = 1500 kg.m-3, μ = 12  10-3 Pa.s
1500  1.5  0.67 2
Hence: Re  = 84169 > 104
0.012
Hence from Table (11.1), we get: KT = 5.75
From equation (11.8), we get:
P  5.75 1.53  0.67 5 1500 = 3930 W.
3930
Hence, motor power = = 6.29  7 hp
0.85  735

Example 11.2
If the vessel of the previous example is used to mix a polymeric compound of viscosity
11.2×105 cP and density = 1200 kg.m-3, what should be the power of the motor used in
that case?
Solution:
1200  1.5  0.67 2
Re  = 6.73 < 10
120
Hence equation (11.7) can be used with KL = 65. We get:
135
 Agitation and Mixing

P  65 1.52  0.67 3 120 = 5278 W

5278
Hence, motor power = = 8.44  9 hp
0.85  735

Example 11.3
The same vessel is now used to mix a monomer of viscosity = 150 cP and density = 1200
kg.m-3. What should be the power of the motor used?
Solution:
1200  1.5  0.67 2
Re  = 5387
150  10 3
This value of Re situate the flow pattern in neither laminar nor turbulent regime so that
we cannot apply either corresponding equation.
Referring to equation (11.5), the given data show that:
0.67 134 2
S1  S 2   0.33 , S 4   0.2 , S 6   1
2 670 2
These values correspond to curve (A) in Figure (11.8). For a value of Re ≈ 5400, we get:
Po ≈ 6. From the definition of Po, we get:
P = Po. .n 3 .De5 = 6 1200 1.53.  0.67 5 = 3280 W.
The motor power will then be:
3280
= 5.25  6 hp
0.85  735

11.4 Mixing of liquids

11.11.1 Mixing of miscible liquids
The previous sections dealt with predicting the power necessary for agitation of liquids.
In case of multiphase mixing, the operation is much more difficult to describe. The
pattern of flow in single phase agitation is usually predictable given a certain mixer
design, but in multiphase mixing the extent to which uniformity is reached often depends
on such factors as color change or reaching uniform temperature. The main factor of
interest in multiphase mixing is to predict the time required to get a uniform mix under
the given operating conditions.
In case of miscible liquids blending is usually performed in relatively small process
vessels by propellers or turbines that are usually centrally mounted. In case of large
quantities to be mixed, the agitator may be idle most of the time and be turned on only to
blend any formed stratified layers of liquid.
The prediction of the time required to get almost complete mixing is based on the
following assumptions:
136
 Agitation and Mixing

 Efficiency of mixing = 99%

 Number of circulation loops = 5
 Re > 2000
Given these assumptions it was possible to establish the following equation for predicting
the required time of mixing for standard six blade turbine:
H .Dt
t M  4.3  (11.9)
n.De2
For lower values of Re, the mixing time is appreciably higher.
For Re < 2000, in case of standard six blade turbines, a friction factor has been defined
by Norwood and Metzner that makes it possible to predict the mixing time over a large
range of Reynolds numbers:
2 1
 De   Dt  2 
1

f  n.t M .  .  .Fr 6 (11.10)

 Dt   H 
The relation between this friction factor and Reynolds number for Re < 2000 takes the
form shown in Figure (11.9)

1000

100
f

10

1
1 10 100 1000 10000

Re

Fig (11.9): Mixing time for miscible liquids in 6 blade turbine baffled vessels

Finally, it is worth mentioning that the power required for mixing can still be calculated
from equations (11.7) for laminar regime (Re < 10) or (11.8) for turbulent regime (Re >
104) or from charts such as those shown in Figure (11.8).

137
 Agitation and Mixing

Example 11.4
An agitated vessel 1.83 m in diameter contains a six blade turbine impeller 0.61 m in
diameter rotating at 80 rpm. It is used to neutralize a solution of NaOH with concentrated
nitric acid. The depth of solution = 1.83 m. How long would it take to complete
neutralization? Estimate the required power.
(Density of solution = 998 kg.m-3, viscosity = 0.98 cP)
Solution:
We have: n = 1.33 rps, De = 0.61 m, ρ = kg.m-3, μ = 0.98  10-3 Pa.s
998  1.33  0.612
Re  = 5  105
0.00098
Since Re > 2000, then equation (11.9) can be used. We get:
1.83  1.83
t M  4.3  ≈ 30 s
1.33  0.612
Since Re > 104, then: P  K T .n 3 .De5 . = 5.75×11.333×0.615×998 = 1140 W

Example 11.5
An agitated vessel 2 m in diameter contains a six blade turbine impeller 0.67 m in
diameter rotating at 60 rpm. It is used to mix two miscible elastomers. The depth of
solution = 2 m. How long would it take for blending to be complete? (Density of solution
= 1100 kg.m-3, viscosity = 500 cP)
Solution:
We have: n = 1 rps, De = 0.67 m, ρ = 1100 kg.m-3, μ = 500  10-3 Pa.s
1100  1 0.67 2
Re  = 987 < 2000
0.5
From Figure (11.9), for Re ≈ 1000, we get f ≈ 9
12  0.67
And Fr = = 0.068
9.81
From equation (11.10)
2 1
 De   Dt  2 
1
 
f  n.t M .  .  .Fr 6 , hence:
 Dt   H 
1
2 1
 0.67   2  2 
9  1  t M .  .  .  0.068 6
 2  2
From which: t ≈ 51 s

138
 Agitation and Mixing

11.11.2 Mixing of immiscible liquids

Various types of equipment are used to disperse one liquid in another immiscible liquid
(e.g. benzene in water). A stirred tank can be commonly used to disperse a liquid in the
form of droplets of size varying from 0.05 to 1 mm. if no agitation takes place these
droplets will either rise or fall in the continuous liquid phase depending on the relative
density of the two liquids.
In a stirred tank, the average drop size will depend on a balance between breakup of large
drops in regions of high shear and coalescence of smaller drops in regions of lower shear.
The applied stress tends to deform the drop while its surface tension will resist such
deformation. To that aim, it is necessary to define an average drop size.
Let the volume of dispersed phase per unit volume of continuous phase = Vd. Hence for N
spherical droplets of mean size Ds, we get:

Vd  N . .Ds3 (11.11)
6
Denoting by Ad the total area of droplets, hence:
Ad  N . .Ds2 (11.12)
In practice, Ad represents the interfacial area per unit volume of continuous phase.
Dividing equation (11.11) by (11.12), we get:
6.Vd
Ds  (11.13)
Ad
So, the mean droplet size can be obtained by determining the volume of the dispersed
phase and the interfacial area.
Since the experimental determination of interfacial area is a difficult task, empirical
correlations have been proposed that relate the mean droplet size Ds to the impeller
diameter De and the volume of dispersed phase per unit volume of continuous phase Vd.
These correlations include a dimensionless group known as the Weber number We
which represents the ratio between the kinetic energy of liquid at impeller tip and its
surface tension. It is defined as follows:
 c .n 2 .De3
We  (11.14)
d
Where, ρc is the density of continuous phase and γd the surface tension of dispersed phase
(N.m-1)
One such correlation applicable for the dispersion of low viscosity liquids in small tanks
is:
Ds
 0.058  ( 1  5.4Vd ).We 0.6 (11.15)
De

139
 Agitation and Mixing

Example 11.6
8% cyclohexane (by volume) is dispersed in water at 25oC in a baffled vessel 300 mm in
diameter with a depth of 350 mm. The agitator is a standard six blade turbine 100 mm in
diameter. The stirrer is run at 360 rpm.
Estimate the power consumption and the mean droplet size.
(Density of cyclohexane = 760 kg.m-3, viscosity ≈ 1 cP, surface tension = 0.046 N.m-1)
Solution:
Mean density of mixture = 0.08  760  0.92 1000 = 980.8 kg.m-3
Speed of revolution = 360 rpm ≡ 6 rps
Hence:
980.8  6  0.12
Re = 3 = 5.88  105 > 1011.
10
So, equation (11.8) can be used to predict the power with KT = 5.75 (Table 11.1):
P  K T .n 3 .De5 . = 5.75  63  0.15  980.8 = 12 W
The mean droplet size is calculated from equation (11.15).
1000  6 2  0.13
We = = 783
0.046
Ds
 0.058  ( 1  5.4  0.08 )  7830.6 , hence: Ds = 11.52 10 4 m ≡ 0.152 mm
0.1

11.5 Scale – up of mixers

One main problem associated with the industrial use of mixers and agitators is to scale –
up laboratory results to industrial scale. The power required for a mixing operation that is
calculated or determined experimentally is usually different from that that would be
required for a larger agitated vessel.
In order to scale – up laboratory results, it has been suggested that some factors or
combinations of factors have to be maintained constant throughout scaling up. This is
known as similarity between small and large scale setups. There are several such
similarities but the two major ones that are more frequently used are the equal tip
velocities similarity and the equal power per unit volume similarity.
In the first case, we assume that the velocities at tip of impeller are similar. Denoting by
(1) and (2) the values of the small scale and large scale setups respectively, we get the
following condition:
n1 .De1  n2 .De 2 (11.16)
In the second case it is assumed that the power per unit volume in both cases is the same.
Now the power can be obtained from the definition of power number where as the

volume of liquid = .Dt2 .H
4
140
 Agitation and Mixing

The power per unit volume is therefore:

P Po  n 3 .De5 .  4.Po .  De   Da  
2

         ( n .De )
3 2

V     Dt   H 
 Dt .H
2

4
The first bracket contains ratios that are usually kept constant whether in small or large
tanks. Also the power number is sensibly the same so that equal power per unit volume
similarity leads to the following equation:
n13 .De21 = n23 .De22 (11.17)
An important condition imposed in scaling – up is that the total number of revolutions to
effect a certain blending should be kept constant. That is:
t M 1  n1  t M 2  n2 (11.18)

Example 11.7
A pilot vessel 300 mm in diameter is agitated by a six blade turbine impeller of diameter
= 100 mm. it is used to mix two miscible liquids of height 300 mm. When the speed of
rotation is 320 rpm, the mixing time was found to be 15 s. Estimate the required mixing
time.
Use the two scaling – up similarities enunciated above.
Solution:
320
Since n1 = = 5.33 rps and De1 = 0.1 m, then assuming the ratio between vessel and
60
impeller diameter to be the same, De2 = 0.6 m.
(1) For equal tip speed similarity: From equation (11.16):
5.33  0.1  n2  0.6 , so that n2 = 0.89 rps
From equation (11.18):
15  5.33  t M 2  0.89
Hence: tM2 ≈ 90 s
(2) For equal power per unit volume: From equation (11.17):
5.333  0.12 = n23  0.6 2 so that n2 = 11.6 rps
From equation (11.18):
15  5.33  t M 2 1.6
Hence: tM2 ≈ 50 s
We note the discrepancy between the results obtained on using different similarity
models. This is frequent in mixers scale – up. It is usually recommended to use the

141
 Agitation and Mixing

longest time obtained from either model or simply to rely on experimental data taken on
the industrial setup.

11.6 Mixing of particulate solids

11.6.2. Introduction
Equipment in which solid materials are mixed may be used for a number of operations.
Blending of ingredients may be the main objective, as, for example, in the preparation of
feeds, insecticides, fertilizer, glass batches, packaged foods, and cosmetics. Other
objectives may include cooling or heating such as in the cooling of limestone or sugar or
the preheating of plastic prior to calendering. Drying or roasting of the solids is
sometimes desired. In some applications, such as polymerization of plastics, catalyst
manufacture, or the preparation of cereal products, the solids mixture may be reacted.
Coating is desired in some cases, as in the manufacture of pigments,dyes, minerals,
candy, and other food products and in the preparation of feeds. In certain of these cases,
small amounts of liquid may beadded, but the end product is a solids mixture. Sometimes
agglomerates are desired, as in the preparation of food products, pharmaceuticals,
detergents, and fertilizer. Often size reduction is desired while solids are being mixed. In
all cases, the mixing of solids occurs. However, in some of these operations, the details of
the equipment to accomplish operations other than pure blending may become a major
problem.

11.6.2. Factors affecting solid mixing

Wide differences among properties such as particle-size distribution, density, shape, and
surface characteristics (such as electrostatic charge) may make blending very difficult. In
fact, the properties of the ingredients dominate the mixing operation. The most
commonly observed characteristics of solids are as follows:
(a) Particle size distribution: This tells the percentages of the material in different size
ranges.
(b) Bulk density: It is not a constant value and can be decreased by aeration and
increased by vibration or mechanical packing.
(c) True (Particle) density
(d) Particle shape: Some particulate solids are pellets, egg shapes, chips, blocks, spheres,
flakes, rods, filaments, crystals, or other irregular shapes.
(e) Surface characteristics: These include surface area and tendency to hold static
charges.
(f) Flow characteristics: The most important being the angle of repose and flowability.
A steeper angle of repose would indicate less flowability. The term “lubricity” has
sometimes been used for solid particles to correspond roughly to viscosity of a fluid.
(g) Friability: This is related to the ease of grinding. Abrasiveness of one solid ingredient
upon another should also be considered.
(h) State of agglomeration: This refers to whether the particles exist independently or
adhere to one another in clusters. The kind and degree of energy employed during

142
 Agitation and Mixing

mixing and the friability of the agglomerates will affect the extent of agglomerate
breakdown and particle dispersion.
(i) Moisture or liquid content of solids: Often a small amount of liquid is added for dust
reduction or special requirements (such as oils for cosmetics). The resultant material
may still have the appearance of a dry solid rather than a paste.
(j) Density, viscosity, and surface tension: These are properties at operating temperature
of any liquid added.
(k)Temperature limitations of ingredients. Any unusual effects due to temperature
changes which might occur (such as heat of reaction) should be noted.

11.6.11. Measuring uniformity of mixing

The uniformity of mixing can be measured by statistical means depending on the number
of particles of each component and their mass fraction. This defines a mixing index
which is 0 for totally segregated powders and 1 for a totally homogeneous mixture.
In practice, however, it is difficult to accurately evaluate the value of this index. Instead,
a defining property (such as true density) can be followed with time by taking several
random samples of the mixture from different locations inside the mixture at regular time
intervals. The standard deviation of each set of values taken at a certain time is defined
by the following equation:

( x i  X )2
 i 1
(11.19)
n 1

Where:
n: Number of samples taken
xi: Individual values of the defining property
X: mean value of xi
The value of standard deviation is followed with time and will tend to stabilize after a
certain time. This practically indicates that a reasonable degree of homogenization has
been reached. To avoid having to scale – up such results it is preferable to effect these
trials on the industrial mixer to be used.

11.6.11. Mixing mechanisms

There are several mechanisms by which solid particles are mixed.
These include small – scale random motion (diffusion), large scale random motion
(convection) and shear. Motions which increase the mobility of individual particles will
promote diffusive mixing. This usually leads in time a great degree of homogeneity.
Diffusive mixing takes place when particles are distributed over a freshly developed
surface and when they are given increased mobility. A plain tumbler gives the former,
while an impact mill gives the latter.
For most rapid mixing, in addition to diffusive (fine-scale) mixing, there should be a
means by which large groups of particles are intermixed.

143
 Agitation and Mixing

This is accomplished by either convective (using a ribbon mixer) or shear mechanism

(using a tumbler mixer).
The diffusion mechanism occurs readily for free-flowing powders in which individual
particles are highly mobile, but is inhibited by cohesion among particles. It follows that
cohesive powders, containing fine material or liquid phases, are relatively difficult to
mix. At the same time, reduced particle mobility inhibits de-mixing so that once mixed,
cohesive powders tend to remain so. Free-flowing powders, on the other hand are prone
to de-mixing during any transport/handling operation. The beneficial effects, noted
above, of liquid addition presumably result from increased cohesion.

11.6.5. Mixing equipment

(a) Tumblers
These are simple drums that rotate about a central axis. They are suitable for gentle
blending and are capable of handling large volumes. They are also easy to clean and are
mostly used for mixing dense powders and abrasive materials. However, they cannot be
used for breaking agglomerates. Figure (11.10) shows two such types: The double cone
blender and the V – type (or twin shell) blender.

(a) (b)

Fig (11.10): (a) Double cone blender

(b) Twin shell blender

Fig (11.11): Double cone baffled mixer

Whenever it is required to break down

agglomerate formation, it is
necessary to provide for some means to this
aim inside the mixer. For example, as can be
seen from Figure (11.11), a double cone
mixer can be internally fitted with baffles, or a ball mill can be used if simultaneous
grinding and mixing is required.
(b) Stationary shell mixers

144
 Agitation and Mixing

There are a number of different types of mixers in which the container is stationary and
material displacement is accomplished by single or multiple rotating inner mixing
devices. One of the most common in this respect is the ribbon mixer, shown in
Figure(11.12). In this type, ribbon cross section and pitch, clearances between outer
ribbon and shell, and number of spirals on the ribbon are some features which can be
varied to accommodate materials ranging from low-density finely divided materials that
aerate rapidly to fibrous or sticky materials that require positive discharge aid. A broad
ribbon can be used for lifting as well as for conveying, while a narrow one will cut
through the material while conveying. The ribbon is adaptable to batch or continuous
mixing.

Fig (11.12): Ribbon mixer Fig (11.13): Muller mixer

Another type is the Muller mixer shown in Figure (11.13). It is particularly useful for
agglomerate breakdown and in case of moist solids. However, it is mostly used for batch
operations.
Whenever continuous operation is required, a commonly
used type is the double shaft mixer, shown in Figure
(11.14). It consists of two shafts with either paddles or
screws encased in a cylindrical shell. It is particularly
suited for mixing of non – free flowing or moist solids.
Besides, it is easily adaptable for either heating or
cooling while mixing. It can also be used to mix pastes of
low water content (< 20%).

Fig (11.14): Double shaft mixer

11.6.6. Power required to operate a tumbling mixer

A tumbler mixer can be approximated to a hollow cylinder that contains a certain amount
of solid particles rotating about a central shaft. The following analysis gives an
approximate method for calculating the power required to operate such mixer.
145
 Agitation and Mixing

Consider Figure (11.15) representing the powder at rest inside the tumbling mixer. The
powder usually fills 30 to 40% of its volume. If the filling ratio = f and the dimensions of
mixer D and L, then the mass of solid inside mixer is:

M = D 2 Lf B (11.20)
4
Where, ρB is the bulk density of solid.

θ
d

Fig (11.15): Tumbler mixer

The theoretical torque required to move the mixer is:

T = I.α (11.21)
The moment of inertia is the sum of that of rotating cylinder and that of the charge I = I1
+ I2 :
I1 = L( R22  R12 )  R12 (11.22)
I2 = M.d2 + 0.00875M.D2 (11.23)
Where: d is the distance between the center of gravity of the cross section of the mixer
and that of charge. The value of d is obtained from:
3 
2 D sin
d= 2 (11.24)
3(  sin  )
The power requirement is obtained from equation:
P = T.ω (11.25)

Example 11.8
A tumbling mixer can be approximated as a cylinder of diameter = 0.8 m and length =
1.2 m. It is used to mix a plastic powder (Sp. Gr. = 0.9) with a pigment (sp. Gr. = 2.8)
added to the powder in a mass ratio of 5:3 respectively. The powder mix has a porosity of

146
 Agitation and Mixing

0.4 and fills 35% of the mixer volume. Calculate the power required to drive the drum at
180 rpm. (Assume steel wall thickness = 8 mm and the time required to reach 180 rpm
from rest = 3 s.)
Solution:
The mean density of the powder mix is calculated from:
8 5 3
  Hence ρ = 1.56 g.cm-3
 2.8 0.9
Bulk density = ρB = 1.56 (1 – 0.4) = 0.936 g.cm-3 ≡ 936 kg.m-3

From equation (11.20): M =  0.8 2  1.2  0.35  936 ≈ 198 kg
4
Calculation of angle α:
The area of the segment in Figure (11.15) = 35% of the area of the circle = 0.275D2
D2
(  sin  ) = 0.275D
2
The area of a segment in a circle of diameter D =
8
Hence θ – sin θ = 2.2
By trial we get θ ≈ 2.66 rad. (≡ 152.4o).
152.4
2  0.8  sin 3
Substituting in equation (11.24): d= 2 = 0.222 m
3  2.2
Mass of drum = L( R  R ) = π×7800×1.2×(0.4082 – 0.42) = 190 kg
2
2 1
2

Moment of inertia = I = MR2 = 190×0.42 = 30.4 kg.m2

Moment of inertia of the rotating solid obtained from equation (11.22) = 198×0.2222 +
0.00875×198×0.82 = 10.87 kg.m2
Total moment of inertia = 30.4 + 10.87 = 41.27 kg.m2
Angular speed ω = 2π×180/60 = 18.85 s-1
Angular acceleration = 12.6 / 3 = 6.28 s-2
Torque = I.α = 41.27×6.28 = 259.2 N.m
Substituting in equation (11.25), we get: P = 259.2×18.85 = 4885 W
Assuming an efficiency of 0.7, we get: P = 6979 W → 10 hp

147