EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

EXPERIMENT 4
SOLIDS HANDLING STUDY EQUIPMENT.

The exercises in this module consist of four separate experiments as follows:

Experiment A: Removal of Dust from a Gas Stream Using a Cyclone

Experiment B: Size Reduction Using a Ball Mill
Experiment C: Screen Analysis Using Sieve Shaker
Experiment D: Determination of Mixing Index of Solids

All the above experiments are related to solids handling which is an important in many
process operations. The experiments are arranged in a single ring as shown in Figure 4.1
below.

Figure 4.1 The Experimental Rig

39
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

The rig consists of a cyclone, ball mill, a screen separator and a mixer.

Application of the rig for different experiments are detailed below.

4.1 Experiment A: Removal of Dust from a Gas Stream Using a Cyclone.

4.1.1 Introduction:

Cyclone is a settling device in which a strong centrifugal force, acting radially, is

basically used in place of a relatively weak gravitational force acting vertically.
Normally, cyclone is used for removing particles with sizes of greater than 10 μm.
Unless the particles sizes are greater then 25 μm, the efficiency of a cyclone very often
does not exceed 90%.
4.1.2 Theory
The collection efficiency η (%) of a cyclone can be calculated from the following
equation:
⎛ mass in - mass out ⎞
η=⎜ ⎟ x 100% (A1)
⎝ mass in ⎠
where, (mass in - mass out) is the mass collected in the collector.
The pressure drop, ΔP, is given by the following equation:
1 2
ΔP = ν H (A2)
2 1 ν
where
KA
Hν = (i.e. the column inlet velocity number)
D e2
ΔP = pressure drop
ν1 = inlet gas velocity
ρ = gas density
A = inlet cross-sectional area
K = cyclone configuration constant
DP = outlet pipe diameter

40
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

Since gas velocity is proportional to the gas volumetric flow rate Q, the pressure drop is
then directly proportional to the square of the velocity. High collection efficiency is
normally related to high pressure drop. Simple cyclone usually has pressure drops within
0.5 to 2.0 in. H2O to 2.0 in. H2O whereas high collection efficiency cyclone usually has
pressure drops between 2.0 to 6.0 in. H2O.

4.1.2 Experimental Procedure

In this experiment students will be demonstrating ejector and cyclone operations.

Students must be familiar with each items included in the cyclone system before
operating. This experiment requires approximately 100 g of particle sample (e.g. food
particles) having size smaller than 22 mesh.
1. Make sure all connections are tight before starting the experiment and there is no
particles left over in the system.
2. Place a collecting pan under sieve with 22 mesh number. Sieve the sample and
determine the density of the sample collected in the pan.
3. Switch on the air compressor and close its valve. The compressor will stop
automatically when it reaches the pre-set pressure.
4. Weigh accurately 5.0 gram of particle sample. Fill the particle feed tube with the
sample. Put the feed tube into the sample beaker.
5. Open the air compressor valve and adjust the flow meter reading to 80 liter/min.
Notice that pressure drop changes with any variation in the gas flow rate. Record
the pressure drop from the manometer.
6. Insert the flexible feed hose into the feed tube so that the particle sample is sucked
out by the ejector and goes into the cyclone.
7. When the feed tube is empty, close the air compressor valve.
8. Determine the weight of the sample collected in the collecting tube, discard the
sample.
9. Before proceeding with a different gas flow rate, remove the sample left over (if
any) from the system. This can be done by passing a high velocity air through the

41
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

system for about 30 seconds. (CAUTION: Very high air velocity might cause
the manometer liquid to shoot out).
10. Repeat steps 4 to 9 for various flow meter readings from 60 to 180 liter/min. at an
interval of 10 liter/min.
Note: If the pressure is too low, wait until it reaches the pre-set pressure before
proceeding with step 5.

4.1.3 Results

Flow meter Reading = 60 to 180 lit./min.

Gas flow rate, = Q lit./min.
Weight of sample in feed tube = WS,total gram
Weight of empty collecting tube = Wc gram
Pressure drop = ΔP mm H2O
Weight of collecting tube + sample = Wc+s gram

4.1.4 Calculations

Weight of sample (in collecting tube) = Wc + s - Wc

From equation (A1),
( Wc + s − Wc )
Efficiency, η = x 100%
WS , total

From equation (A2),

2 ΔP 2ΔP D P2
Velocity, ν1 = =
Hν KA

4.1.5 Report
1. Complete all the data in Table 1.
2. Record ambient temperature (°C) and pressure (atm.)
3. Plot (a) (ν1)2 versus experimental ΔP and
(b) Q versus η

42
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

4.2 Experiment B: Size Reduction Using a Ball Mill

4. 2.1 Introduction:
The term size reduction is applied to all the methods by which particles of solids are
disintegrated (cut or broken) into smaller pieces. Solids may be broken in many
different ways, but only four of them are commonly used in size-reduction machines.
They are (1) compression, (2) impact, (3) attrition, or rubbing, and (4) cutting. Size
reduction equipment is divided into crushers, grinders, ultra fine grinders, and cutting
machines. The system is equipped with a batch type cylindrical ball mill which falls
under the grinders category. In a ball mill most of the reduction is done by impact as the
balls drop from near the top of the shell. The load of balls in a ball mill is normally such
that when the mill is stopped, the balls occupy about one-half the volume of the mill.
The void fraction in the mass of balls at rest is typically 0.40.

When the mill is rotated, the balls are picked up by the mill wall and carried nearly to the
top, where they break contact with the wall and fall to the bottom to be picked up again.
Centrifugal force keeps the ball in contact with the wall and with each other during the
upward movement. Most of the grinding occurs at the zone of impact, where the free-
falling balls strike the bottom of the mill. The faster the mill is rotated, the farther the
balls are carried up inside the mill and the greater the power consumption. If the speed is
too high, however, the balls are carried over and over the mill is said to be centrifuging
and little or no grinding is done. The speed at which centrifuging occurs is called the
critical speed. Therefore, operating speeds must be less than the critical.

4.2.3 Theory
There are two forces acting on the ball as shown in Figure 4.1. The first is the force of
gravity mg/gc, where m is the mass of the ball. The second is the centrifugal force (R - r)
ω2/gc, where ω = 2πn and n is the rotational speed. The centrifugal component of the
force of gravity (mg/gc) cos α opposes the centrifugal force. As long as the centrifugal
force exceeds the centripetal force, the particle will not break contact with the wall. As
the angle α decreases, however, the centripetal force increases, and unless the speed

43
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

exceeds the critical, a point is reached where the opposing forces are equal and the
particle is ready to fall away. The angle at which this occurs is found by equating the two
forces, giving

g
m cos α =
[
m 4π 2 n 2 ( R − r ) ]
gc gc
(B1)

Figure 4.1 : Forces on ball in ball mill

At the critical speed, α = 0, cos α = 1, and n becomes the critical speed nc. Then
1 g
nc = (B2)
2π R−r
Ball mills are usually run at 65 to 80 percent of the critical speed. The maximum amount
of energy that can be delivered to the solid being reduced can be computed from the mass
of the grinding medium, the speed of rotation, and the maximum distance of fall. In an
actual mill the useful energy is much smaller than this, and the total mechanical energy
supplied to the mill is much greater. As the product becomes finer, the capacity of a
given mill diminishes and the energy requirement increases.

44
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

4.2.3 Experimental Procedure:

In this experiment students will be demonstrating a batch cylindrical ball mill operations
in which coarse salt will be reduced to powder. Students must be familiar with the ball
mill before operating. Approximately 2 kg of coarse salt (16 mesh size or larger) is
required for this experiment. Student will carry out simple screen analysis for each
operation at different critical speed.
1. Make sure that the ball mill is clean and properly assembled. Cleaning the ball
mill and filling the mill with grinding balls must be done by a well trained
technician or authorized personnel.
2. Calculate the critical speed nc of the ball mill.
3. Weigh accurately 300 g of coarse salt (16 mesh size or larger) and pour into the
ball mill through the inlet. Secure the inlet cap tightly.
4. Set the speed control knob to about 45% of the critical speed (nc) calculated in
step 2. Record time to.
5. Switch on the ball mill and let it run for 15 minutes.
6. Place a collecting tray underneath the ball mill. Stop the ball mill and remove the
inlet cap and remove the sample. Keep the powder asid e for screen analysis later on.
7. Repeat steps 3 to 6 at different percentage of critical speed: 70%, 100% and
130%.

4.2.4 Results:

a) Critical speed calculation.

acceleration of gravity = g = 980 cms-2
radius of ball mill = R = ____ cm
Grinding Media Specific Gravity r (in.) nc (rpm)
Zirconia 5.5 0.250
Stainless Steel Ball 0.375

45
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

4. 3. Experiment : Screen Analysis Using Sieve Shaker

4.3.1 Introduction
A collection of particles of different sizes could be classified into different grades by
means of a set of standard screen is arranged serially in a stack, with the smallest mesh at
the bottom and the largest at the top. The sample is placed on the top screen and the
stack shaken mechanically for a definite time. The particles retained on each screen are
removed and weighed, and the masses of the individual screen increments are converted
to mass fraction or mass percentages of the total sample. Any particles that pass the
finest screen are caught in a pan at the bottom of the stack.
In order to specify the size range of an increment, two mesh numbers are needed in a
form of “A/B”. Thus, 16/22 means “through 16 mesh and on 22 mesh”. Table C1 shows
a typical screen analysis where i is the number of the screen starting at the bottom of the
stack, i.e., i = 1 for the pan. The symbol Dpi means the particle diameter equal to the
mesh opening of screen i.

Table C1: Screen Analysis

Mesh Screen opening Mass fraction Average dia. in Cumulative fraction
Dpi, mm retained, x1 increment, Dpi , mm smaller than Dpi

4 4.699 0.0000 - 1.0000

6 3.327 0.0251 4.013 0.9749
-
-
200 0.074 0.0031 0.089 0.0075
Pan - 0.0075 0.037 0.0000

A fair picture of particle size distribution is given by a plot of x1/(Dpi + 1 - Dpi), where
(Dpi + 1 - Dpi), is the particle size range in increment I. Cumulative plots are made by
plotting cumulative fraction smaller than Dpi versus particle diameter.
4.3.2 Theory:

46
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

The individual solid particles are also characterized by their size, shape and density. The
shape of an individual particle is conveniently expressed in terms of the sphericity Φs,
which is independent of particle size. For a spherical particle of diameter Dp, Φs = 1; for
a non-spherical particle, the sphericity is defined by the relation
6νp
Φs = (C1)
D P sp

where,
Dp = equivalent diameter or nominal diameter of particle
sp = surface area of one particle
νp = volume of one particle.
In general, diameters may be specified for any equi-dimensional particle. Particles which
are not equi-dimensional, i.e., which are longer in one direction than in others, are often
characterized by the second longest major dimension. For needle like particles, for
example, Dp would refer to the thickness of the particles, not their length. By
convention, particle sizes are expressed in different units depending on the size range
involved. Coarse particles are measured in inches or millimeters; fine particles in terms
of screen size; very fine particles in micrometers or manometers. Ultra-fine particles are
sometimes described in terms of their surface area per unit mass, usually in square meters
per gram.

In a sample of uniform particles of diameter Dp, the total volume of the particles is m/ρp,
where m and ρp are the total mass of the sample and the density of the particles,
respectively. Since the volume of one particle is νp, the number of particles, N in the
sample is
m
N= (C2)
ρpνp

The total surface are of the particles is, from Eqs. (C1) and (C2),
6m
A = N x sp = (C3)
Φs ρp D p

47
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

In differential analysis information from a particle-size analysis is tabulated to show the

mass or number fraction in each size increment as a function of the average particle size
in the increment (Figure 4.2a). The dashed line is used to approximate the distribution.
Another way to present the information is through a cumulative analysis obtained by
adding, consecutively, the individual increments, starting with that containing the
smallest particles, and tabulating or plotting the cumulative sums against the maximum
particle diameter in the increment (Figure 4.2b). In a cumulative analysis the data may
appropriately be represented by a continuous curve.

Figure 4.2 : Particle-size distribution for powder

(a) differential analysis
(b) cumulative analysis

If the particle density ρw and sphericity Φs are known and they are constant, the specific
surface Aw (i.e. the total surface area of a unit mass of particle) is given by
6x 1 6x 2 6x n
Aw = + + ......+
Φs ρp Dp1 Φs ρp Dp 2 Φs ρp Dpn
(C4)
6 n x
= ∑ 1
Φs ρp t = 1 Dpt

where,
subscripts = individual increments

48
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

x1 = mass fraction in a given increment

n = number of increment
Dpt = average particle diameter (taken as arithmetic

average of smallest and largest particle

diameters in increment)
The volume-surface mean diameter Ds , which is related to Aw, is defined by
6
Ds = (C5)
Φs A w ρp

Substituting Equation (C4) into Equation (C5) gives

1
Ds = n (C6)
(
∑ x 1 / Dpt
t=1
)
The arithmetic mean diameter DN is

( ) ( )
n n
∑ N 1 Dpt ∑ N t Dpt
t=1 t=1
DN = n
= (C7)
NT
∑ N1
t=1
where NT is the number of particles in the entire sample.
The mass mean diameter D w , is found from the equation

( )
n
D w = ∑ x 1 Dpt (C8)
t=1

The volume mean diameter D v , is given by

1
⎡ ⎤3
⎢ 1 ⎥
⎢
Dv = n ⎥ (C9)
⎢ ∑ 3 ⎥
( x / Dpt )
⎢⎣ t = 1 1 ⎥⎦

For a given particle shape, the volume of any particle is proportional to its “diameter”
cube, or

( )3
νp = a D p (C10)

where a is the volume shape factor. Assuming that a is independent of size, then the total
population in one mass unit of sample Nw is given by

49
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

1 n xt 1
Nw = ∑ = (C11)
aρp t = 1 Dpt aρp D v3
3

4.3.3 Experimental Procedure:

In this experiment students will be demonstrating a sieve shaker operations and then
carry out a screen analysis. Students must be familiar with the operations of the shaker.

1. Read the Operating Manual on the operations of the shaker.

2. Arrange the sieves in the stack with the smallest mesh (i.e. the biggest mesh
number) at the bottom, above the pan.
3. Pour the sample into the top most sieve and spread evenly. Secure the sieve cover
provided.
4. Set the timer to 20 minutes. Switch on the shaker and wait until it stops
automatically.
5. Carefully weigh the sand retained in each screen.

4.3.4 Results:

a) Reconstruct and fill up the table below:

Mesh Screen opening Mass fraction Average dia. in Cumulative fraction
Dpi, mm retained, x1 increment, Dpi , mm smaller than Dpi

8 2.000
16 1.000
22 0.710
30 0.500
60 0.250
100 0.150

Pan

50
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

b) Plot mass fraction, x1 vs. particle size, Dp, (mm)

c) Plot mass fraction smaller than Dp vs. particle size, Dp, (mm)

4.4 Experiment D: Determination of Mixing Index of Solids

4. 4.1 Introduction:

Mixing is an important operation in nearly all chemical processes. Mixers can be

categorized into mixer for pastes and mixer for dry solids. The Lasera LS-SHE 400 is
equipped with a V-cone blender which is suitable for mixing dry solids. During
operations, the V-cone blender is partly filled with two different materials and rotated
about a horizontal axis for 10 to 20 min.
4.4.2 Theory
The effectiveness of a solids blender is measured by a statistical procedure. Spot samples
are taken at random form the mix and analyzed. The standard deviation of the analyses s
about their average value x is estimated from

(D1)
With granular solids the mixing index is based on the standard deviation that would be
observed with a completely random, fully blended mixture.
Consider a completely blended mixture of salt and sand grains from which N spot
samples, each containing n particles, are taken. Suppose the fraction of sand in each spot
sample is determined by counting particles of each kind. Let the overall fraction, by
number of particles, of sand in the total mix be μp. If n is small (say about 100), the
measured fraction x1 of sand in each sample will not always be the same, even when the
mix is completely and perfectly blended; there is always some chance that a sample
drawn from a random mixture will contain a larger (or smaller) proportion of one kind of
particle than the population from which it is taken. Thus for any given size of spot

51
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

sample there is a theoretical standard deviation for a completely random mixture. This
standard deviation σe is given by

σe =
(
μp 1 − μp ) (D2)
n
For granular solids the mixing index Is is defined as σe/s. From Eqs. (D1) and (D2) Is
becomes

Is =
σe
=
( )
μ p 1 − μ p ( N − 1)
(D3)
i = 1( x1 − x)
s 2
n∑ N

Typical results for blending salt crystals (NaCI) with Ottawa sand in a small tumbling
blender are shown in figure 4.4. For the first 40 min. the mixing index 1s rose from a
very low value to about 0.7; it fluctuated for a time between 0.55 and 0.7 and then, at
very long mixing times, began to fall steadily. Mixing is initially rapid, but in this type of
mixer it is never perfect. The ingredients of the mix are never blended in a completely
random way. Unbending forces, usually electrostatic, are always at work in a dry solids
blender, and their effects are especially noticeable here. These forces often prevent the
mix from becoming completely blended; when the mixing time is long, they may be, as
shown by figure 4.3, lead to a considerable degree of un-mixing and segregation.

52
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

Figure 4.3 : Blending salt and sand in a tumbling blender

The equilibrium standard deviation for complete mixing σe is used as a reference with
granular solids. The mixing index at zero time (i.e. before mixing begins) is
σ 1
I s, o = e = (D4)
σo n
In mixing, the rate is proportional to the driving force. The mixing index Is is a measure
of how far mixing has proceeded toward equilibrium. It has been found that for short
mixing times the rate of change of Is is directly proportional to 1 - Is or
dl s
= k (1 − I s ) (D5)
dt
where k is a constant. The equilibrium value of Is is 1; therefore the driving force for
mixing at any time can be considered to be 1 - Is. With rearranging and integrating
between limits, Eq. (D5) becomes
1 Is dI s
∫ot dt = ∫I (D6)
k s, o 1 − I s
from which
1 1 − I s, o
t= ln (D7)
k 1 − Is
Substituting form Eq. (D4) gives

53
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

1 1− 1/ n
t= ln (F8)
k 1 − Is
Equation (D8) can be used to calculate the time required for any desired degree of
mixing, provided k is known and unbending forces are not active.

4.4.3 Experimental Procedure:

In this experiment student will be demonstrating the mixing capability of a V-cone mixer.
This experiment requires 1 kg of solids A (e.g. Soya bean) and 1 kg of solids B (e.g.
green bean). Student must familiarize themselves with the operations of the mixer.
1. Place 50-100 g of solids A into one side of the mixer and same amount of solids B
into the other side of the mixer.
2. Set the mixing speed on the control panel.Switch on the mixer and record time, to.
3. Let the mixing run for 15 seconds and stop the mixer. Using the scoop provided,
collect five samples from one side of the mixer and put in separate container.
4. Repeat step 3 several times until complete mixing is seemed to be achieved.
5. Calculate the fraction x1 for each set of sample collected and determine the
mixing index Is.
6. Repeat steps 1 to 5 for different mixing speeds (knob settings 10, 30, 50 etc.).

4.4.4 Results:
a) Reconstruct and complete the table below.

Mixing Time
Mixing t1 t2 t3 ...... t complete mixing
Speed 15 sec. 30 sec. 45 sec. _____ min.
x1 = ____ x1 = ____ x1 = ____ x1 = ____
x2 = ____ x2 = ____ x2 = ____ x2 = ____
Speed 1 x3 = ____ x3 = ____ x3 = ____ x3 = ____
x4 = ____ x4 = ____ x4 = ____ ...... x4 = ____
x5 = ____ x5 = ____ x5 = ____ x5 = ____

54
EKC 493 Chemical Engineering Laboratory III (Unit Operation Lab)

___ rpm x = ____ x = ____ x = ____ x = ____

Is = ____ Is = ____ Is = ____ Is = ____

x1 = ____ x1 = ____ x1 = ____ x1 = ____

x2 = ____ x2 = ____ x2 = ____ x2 = ____
Speed 2 x3 = ____ x3 = ____ x3 = ____ x3 = ____
x4 = ____ x4 = ____ x4 = ____ ...... x4 = ____
x5 = ____ x5 = ____ x5 = ____ x5 = ____
___ rpm x = ____ x = ____ x = ____ x = ____
Is = ____ Is = ____ Is = ____ Is = ____

x1 = ____ x1 = ____ x1 = ____ x1 = ____

Speed 3 x2 = ____ x2 = ____ x2 = ____ x2 = ____
x3 = ____ x3 = ____ x3 = ____ x3 = ____
x4 = ____ x4 = ____ x4 = ____ ...... x4 = ____
___ rpm x5 = ____ x5 = ____ x5 = ____ x5 = ____
x = ____ x = ____ x = ____ x = ____
Is = ____ Is = ____ Is = ____ Is = ____

b) Plot mixing index, Is vs. mixing time, t for each mixing speed.
c) Determine the effective mixing speed.

55