21

Material Degradation Problems

1 INTRODUCTION

Many materials that have to be conveyed are friable, and particles are liable to be
broken when they impact against retaining surfaces, such as bends in the pipeline.
As a consequence there is often a reluctance to use pneumatic conveying systems
for this category of materials, particularly if the material has to be conveyed in
dilute phase and hence at high velocity. There are, however, numerous means by
which the problem can be reduced to an acceptable level.

1.1 Breakage Mechanisms

In some bulk solids handling processes intentional breakdown of the material is
required, as in crushing, grinding and comminution. In many handling and storage
situations, however, unintentional breakage occurs. This is usually termed degra-
dation or attrition, depending on the mechanism of particle breakage. Bulk materi-
als, when pneumatically conveyed, will impact against bends in the pipeline, and
there may be a significant amount of particle to particle interaction in addition.
There may also be frequent impacts against the pipeline walls, and in dense phase
flows particles will slide along the pipeline walls. These collisions and interactions
will produce forces on the particles that may lead to their breakage.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

604 Chapter 21

If particle breakdown occurs readily the bulk solid is said to be friable. Ten-
dency to particle breakdown covers three main situations. The first is a tendency to
shatter or degrade when the bulk solid is subject to impaction or compressive load-
ing. The second is the tendency for fines and small pieces to be worn away by
attrition when bulk solids either rub against each other or against some surface,
such as a pipeline wall or bend. The third is the tendency for materials such as
nylons and polymers to form angel hairs when conveyed, as a result of micro-
melting occurring due to the particles sliding against pipeline walls.

1.2 Magnitude of Problem

Of all conveying systems, dilute phase probably results in more material degrada-
tion and attrition than any other. This is because particle velocity is a major vari-
able in the problem and, in dilute phase conveying, high velocities have to be
maintained. The potential influence of a pneumatic conveying system on a mate-
rial is demonstrated in Figures 21.1 and 21.2 [1], This is a consequence of convey-
ing a friable material at an excessively high velocity in dilute phase suspension
flow in a conveying system with a large number of small radius bends.
Figure 21.1 shows the influence on the cumulative particle size distribution
for the material before and after conveying. The mean particle size, based on the
50% value, has changed from about 177 to 152 urn. The really significant effect,
however, is shown in the fractional size distribution plot in Figure 21.2. In this
magnified plot the effect of degradation on the material can be clearly seen. A
considerable number of fines are produced and even on a percentage mass basis
these cause a significant secondary peak in the particle size distribution.

100
Material before
80 conveying

60
Material after
3 40 conveying

20

40 80 120 160 240

Particle Size - urn

Figure 21.1 Possible influence of pneumatic conveying on cumulative particle size

distribution of a friable material.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 605

3
on
40 Material before
c
C3 conveying
os
§ Particle size
c/5
u 30
Material after
conveying

•S 20

a 10

40 80 120 160 200 240

Particle Size - u.m

Figure 21.2 Possible influence of pneumatic conveying on fractional size distribution

of a friable material.

1.3 Operating Problems

Particle degradation can cause problems in a number of areas on account of the
changes in particle shape and particle size distribution that can result. It is a par-
ticular problem with chemical materials that are coated, for it is the coating that is
generally the friable element of the resulting material. Plant operating difficulties
are often experienced because of the fines produced, and problems in handling
operations can also result after the material has been conveyed.
Apart from the obvious problems of quality control with friable materials,
changes in particle shape can also lead to subsequent process difficulties with cer-
tain materials. The appearance of the material may also change so that it is not so
readily sold. Changes in particle size distribution can affect flow characteristics,
which in the extreme, can change a free-flowing material into one which will only
handle with great difficulty and, with materials for subsequent sale, this can lead to
customer problems.

1.3.1 Filtration Problems

In pneumatic conveying systems plant, operating difficulties can result if degrada-
tion causes a large percentage of fines to be produced, particularly if the filtration
equipment is not capable of handling the fines satisfactorily. Filter cloths and
screens will rapidly block if they have to cope with unexpectedly high flow rates
of fine powder. The net result is that there is usually an increase in pressure drop

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

606 Chapter 21

across the filter, and this could be a significant proportion of the total pressure
available in a low pressure system.
This means that the pressure drop available for conveying the material will
be reduced, which in turn means that the mass flow rate of the material will proba-
bly have to be reduced in order to compensate. If this is not done there will be the
risk of blocking the pipeline. Alternatively, if the filtration plant is correctly speci-
fied, with material degradation taken into account, it is likely to cost very much
more as a result. This, therefore, provides a direct financial incentive to ensure that
particle degradation is minimized., even if it does not represent a problem with
respect to the material itself [2],

1.3.2 Flow Problems

In many systems there is a need to store the conveyed material in a hopper or silo.
Flow functions can be determined for bulk particulate materials, from which hop-
per wall angles and opening sizes can be evaluated, to ensure that the material
flows reliably at the rate required. A change in particle size distribution of a mate-
rial, as a result of conveying operations, however, can result in a significant
change in flow properties. Thus a hopper designed for a material in the "as re-
ceived" condition may be totally unsuitable for the material after it has been con-
veyed. As a result it may be necessary to fit an expensive flow aid to the hopper to
recover the situation.

1.3.3 Potential Explosion Problems

Many materials, in a dust cloud, can ignite and cause an explosion. Dust clouds
are clearly quite impossible to avoid somewhere in a pneumatic conveying system,
and so this poses a problem with regard to the safe operation of such systems. Of
those materials that are explosive, research has shown that it is only the fraction
with a particle size less than about 200 u.m that poses the problem. Degradation
and attrition caused by pneumatic conveying, however, can result in the generation
of a considerable number of fines, particularly if the material is friable. Even if the
material did not represent a problem with respect to explosions in the "as re-
ceived" condition, the situation could be very different after the material has been
conveyed.

1.4 Test Rigs and Data Sources

Little data is available on the degradation of materials in pneumatic conveying
systems. This is partly due to the complexity of obtaining and analyzing the data,
but mainly to the fact that so many variables are involved, together with the prob-
lem of relating the data from one material and situation to another. A particular
problem with data obtained from a pneumatic conveying system pipeline is that it
is very difficult to separate the individual contributions made by the bends and the
straight pipeline.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 607

A further problem is that in a pipeline there is a gradual expansion of the

conveying air, which means that the particle velocity is constantly changing. Ve-
locity is a major variable in particle degradation and so this makes attempts at de-
vising experimental plans and analysis of results very difficult.
The major source of information is probably from the basic research that has
been undertaken with small bench scale test rigs in which particles have been im-
pacted against test materials under controlled conditions. This work has often been
carried out to assist in an understanding of erosive wear problems. Although much
of this work cannot be related directly to pneumatic conveying situations, it can
provide valuable information of a comparative nature on a number of variables in
the process.

1.4.1 Acceleration Tube Device

Test facilities employed are very similar to those used in erosive wear research,
such as whirling arm and acceleration tube devices. A device used by Salman et al
[3] is shown in Figure 21.3 and consists of a linear air gun. One particle was tested
at a time. Compressed air was used to accelerate the particles, and particle velocity
could be varied by adjusting the air pressure.
A cage was used to collect the particles and fragments after impact. The par-
ticle impact velocity was determined by measuring the time required for a particle
to travel from the end of the barrel to the target. A photodiode was located at the
end of the barrel and a loudspeaker was mounted behind the target.

Pressure Cage Particle Impact

Regulating .Angle
Valve Speaker
\

Compressed
Air Supply

Computer High-Speed
Electronic Timer

Figure 21.3 Schematic arrangement of acceleration tube test apparatus and measuring
system for particle impact studies.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

608 Chapter 21

In order to study the particle degradation process, brittle materials were used
to ensure that no plastic deformation should take place. Three types of particle
were used and tested. These were aluminum oxide, polystyrene and glass, and all
the particles were spherical. The majority of the work was carried out with 0-2 in
diameter aluminum oxide particles, with particle velocities up to about 6000
ft/min. For every test, 100 particles were impacted, and the number of unbroken
particles was counted to provide an assessment of the degradation.

2 INFLUENCE OF VARIABLES

The variables in particle degradation are similar to those associated with erosive
wear. Velocity, once again, is probably the most important, but particle size and
concentration also play a part. Particle impact angle is equally important, and has a
major influence with respect to the selection of pipeline bend geometry. The influ-
ence of both particle materials and surface materials must also be given due con-
sideration. As with erosive wear, much of the research work into the subject has
been carried out for various other purposes, and so the range of parameters inves-
tigated is often beyond those associated with pneumatic conveying, but it does
provide useful information on the general trends of the variables.

2.1 Velocity
The relative velocity between particles and surfaces has a major influence on the
nature and extent of the degradation and is probably the most important variable in
the problem. In any collision the kinetic energy of the particles has to be absorbed
and may provide sufficient energy for fracture. If the collision is elastic, with a
high coefficient of restitution, much of the kinetic energy will reappear as particle
velocity. In plastic collisions much of the kinetic energy will be converted to heat.
Low velocity impacts tend to knock small chips from the edges of particles,
whereas high velocity collisions are more likely to shatter particles. In general the
rate of damage has been found to be a power law function of velocity, in much the
same way as the erosive wear process. The range in value of the power coefficient
is also large, and can vary between one and five, depending upon the conveyed
material and the system being considered. The possibility of there being a thresh-
old value of velocity, below which no degradation occurs, is also a possibility.

2. /. / Peas
Agricultural products have been widely used in test work. Segler [4] investigated
the effects of air velocity, moisture content, pipeline diameter and material con-
centration on the damage of peas, as a result of pneumatic conveying. His test loop
was 240 ft long, 4'/2 in bore and contained 4 bends. The results of his tests on the
effect of aii- velocity are presented in Figure 21.4. These showed that the damage
increased approximately with the cube of air velocity.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 609

12

IX
o
^H

S 4

2000 4000 6000

Conveying Air Velocity - ft/min

Figure 21.4 The influence of air velocity on the breakage of peas.

2.7.2 Quartz
Tilly et al [5] carried out impact studies with quartz particles against an alloy steel
target in a rotating arm test rig. They found that the particles incurred a substantial
degree of fragmentation which was dependent upon the velocity of impact. Their
results are presented in Figure 21.5.

100
Particle size - 125-150 urn
80 Taraet Material - 11% Cr steel

60

I
i-M
40

I 20
g?
o
0
0 20,000 40,000 60,000

Particle Velocity - ft/min

Figure 21.5 The influence of particle velocity on the degradation of quartz particles.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

610 Chapter 21

The velocity range comes as a result of their work being applied to aircraft
engines. From this it would appear that for fragmentation to occur it is necessary
to exceed a threshold value of velocity. Below this velocity the particles may be
considered to behave elastically. From Figure 21.5 this would appear to occur at a
velocity of about 3000 ft/min for this material.
In work by Tilly and Sage [6] they impacted quartz particles in the size
range of 100 to 225 urn at velocities of 12,000, 26,000 and 60,000 ft/min. Their
results, in terms of particle size distribution, are presented in Figure 21.6. Al-
though this data is for conveying velocities very much higher than those that
would be encountered in a pneumatic conveying system, they relate to a single
impact and so help to illustrate the nature of the problem, for many materials that
are conveyed are significantly more friable than quartz.

2.7.3 A lum inum Oxide

The results of a program of tests carried out with 0-2 inch aluminum oxide parti-
cles impacted at 90° against a steel target are presented in Figure 21.7 [3"|. In this
plot the experimental data has been included to show how the relationship was
derived and to show the limits of scatter in the results. The relationship is typical
of the results obtained and so where families of curves are presented in subsequent
figures from this program of work, experimental data has been omitted for clarity.
It will be seen from Figure 21.7 that there is a very rapid transition in parti-
cle velocity from zero breakage to total degradation. Below a particle velocity of
about 1800 ft/min only elastic deformation occurs and there is no particle degrada-
tion. Above a particle velocity of about 5000 ft/min, however, the stress induced
by the impact is always sufficient to damage every particle.

100 Particle Velocity

- ft/min

80

60
Before
Impact
40

20 Particles - Quartz
Target - 11% Cr Steel

50 100 150 200

Mean Particle size

Figure 21.6 Influence of particle velocity on size distribution generated with quartz.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 611

100
Particle size = 0-2 in
80 Impact angle = 90°
Target material = Steel

60
C
D 40

X 20

2000 3000 4000 5000 6000

Particle Velocity - ft'min

Figure 21.7 The influence of particle velocity on the degradation of aluminum oxide
particles.

It is interesting to note that within the transition region the number of unbro-
ken particles at any given velocity is very consistent and that a smooth transition is
obtained from one extreme to the other over this range of velocity. This was
probably one of the first research programs to focus on particle degradation in the
velocity range appropriate to pneumatic conveying.

2.2 Particle Size

Tilly et al [5] carried out impact studies with quartz particles against an alloy steel
target in a rotating arm test rig. They found that the particles incurred a substantial
degree of fragmentation which was dependent upon the initial particle size. Their
results are presented in Figure 21.8.
From this it would appear that for fragmentation to occur it is necessary for
the particles to exceed a threshold size of about 10 urn. Below this size the parti-
cles probably behave elastically, for in their test rig the particles would have im-
pacted the target since the tests were carried out in a vacuum.
The results of tests carried out with three different sizes of spherical alumi-
num oxide particles are shown in Figure 21.9 [3], The data for the 0-2 in particles,
which was the reference material in the work, was presented above in Figure 21.7.
Results from similar tests with 0-12 and 0-28 in aluminum oxide particles, also
impacted at 90° against the same steel target are additionally presented in Figure
21.9.
A very significant particle size effect is shown. As the particle size in-
creases, the maximum velocity at which no degradation occurs decreases. The
transition from no degradation to total degradation also changes, with the transi-
tion occurring over a narrower velocity range with increase in particle size.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

612 Chapter 21

100

80

60
S Particle Velocity - 60,000 ft/min
01)
S Target Material - 11% Cr Steel
40

ob 20

50 100 150 200

Initial Mean Particle size - u.m

Figure 21.8 The influence of initial particle size on the degradation of quartz particles.

2.2.7 Particle Velocity Influence

In more recent work on the influence of particle size, fertilizer particles, also hav-
ing particle diameters of 0-12, 0-20 and 0-28 in, were pneumatically conveyed in a
test facility to assess their degradation [7]. In this case the velocity used was that of
the conveying air and not that of the particles. In terms of air velocity the 0-12 in
particles degraded the most and the 0-28 in particles the least.

100

0-
g 60 Impact Angle
^
XJ
-90°
c
40 Target Material
o - Steel
!-~
O
X)
20
0-12

1000 2000 3000 4000 5000

Particle Velocity - ft / min

Figure 21.9 The influence of particle velocity and particle size on the degradation of
aluminum oxide particles.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 613

The reason for this is that when it is the air velocity that is held constant, the
smaller particles are accelerated to a higher velocity than the larger particles, as
illustrated earlier with Figure 15.10. It is because particle velocity has a greater
influence on degradation than particle size that a reversal in the influence of parti-
cle size has occurred.

2.3 Surface Material

With erosive wear of surface materials it has been found that the resilience of the
surface material can have a significant influence on erosive wear, and that rubbers
and polymers can offer better wear resistance than metals having a very high
hardness value in certain cases. Since the mechanisms of erosion and degradation
have many similarities, it is quite possible that resilient materials could offer very
good resistance to particle degradation.

2.3.1 Material Type

Further work by Tilly and Sage [6] showed that fragmentation is also dependent
upon the type of target material. Figure 21.10 shows a comparison of their results
for quartz impacted against nylon and fiberglass, which with their earlier results
for alloy steel demonstrates the complex nature of the problem. Degradation in
terms of the influence of initial particle size is used for the comparison in this case.
The results of tests carried out on four different target materials are pre-
sented in Figure 21.11 [3]. In each case the targets were 0-2 in thick and they were
impacted by 0-2 in diameter aluminum oxide particles at an angle of 90°.

100
xo
Target Material Particle,Velocity
o- /
Impact Angle - 90°
c 80 Steel - 60,000 ft/min

60
I
5h
cd

'•o 40
<L>
a
Q 20
Fiberglass - 50,000 ft/min

0
50 100 150 200 250
Initial Mean Particle size -

Figure 21.10 The influence of initial particle size and target material on the degrada-
tion of quartz particles.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 614 Chapter 21

100 r •*—- •"•— —"•= — Plexiglass

80L ^ ^ . . .
Aluminum
Particle Size - 0-2 in
60
Inroad Angle - 90°
5 40
<+-<
o
I 20
J:

1000 2000 3000 4000 5000

Particle Velocity - ft/min

Figure 21.11 The influence of particle velocity and target material on the degradation
of aluminum oxide particles.

This also shows very clearly that target material can have a very marked ef-
fect on degradation. Although there is little difference in the maximum value of
particle velocity at which no degradation occurs, varying from 2000 ft/min for
steel to about 3000 ft/min for Plexiglas and aluminum, very significant differences
exist in the transition region between no degradation and total degradation. In the
case of the steel and glass targets the transition is very rapid. For the aluminum
and Plexiglas, however, the transition is very slow, and so a high velocity impact
against these materials would only result in limited damage occurring.

2.5.2 Surface Thickness

A similar program to that reported in relation to Figure 21.11 was carried out with
steel targets of varying thickness [7]. If the conveyed material is not abrasive, in
addition, a thin walled surface would also help reduce degradation, for the work
showed a significant reduction in degradation of the particles with an 0-04 in thick
target as compared with an 0'08 in thick target.
The force acting on a particle is equal to its mass times the rate of decelera-
tion. This force must be reduced in order to reduce the damage to particles on im-
pact against a surface. This can be achieved to a certain extent by using either a
resilient surface material or a surface material that will flex on impact.

2.4 Particulate Material

In Figure 21.12 the data for the aluminum oxide of Salman et al [3J is presented
again, together with the results from identical tests carried out with polystyrene
and glass particles. It will be seen from this that polystyrene particles suffer a simi-
lar transition from zero breakage to total degradation.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 615

100

Impact Angle - 90°

^5
o
60 Target Material - Steel

40
Aluminum11
Oxide
1 20

2000 3000 4000 5000 6000

Particle Velocity - ft/min

Figure 21.12 The influence of particle velocity on the degradation of various particu-
late materials.

The only difference is that the transition occurs at a slightly higher velocity
range than the aluminum oxide. That different particulate materials can respond in
totally different ways is clearly demonstrated by the glass particles. No damage
was observed to any of the particles tested up to the maximum particle velocity
investigated of 6000 ft/min.

2.5 Particle Impact Angle

Particle impact angle, oc, was defined on the sketch of the acceleration tube test
device shown earlier in Figure 21.3, and is the same as that used in erosive wear
work. Impact angle has been shown to be a major variable with regard to the ero-
sive wear of surface materials, and hence is an important consideration in terms of
material selection and the specification of components such as pipeline bends. In
relation to particle degradation it is equally important, for as the impact angle re-
duces, so the normal component of velocity decreases [8]. This will have a direct
bearing on the deceleration force on the particles, as discussed above in relation to
Surface Thickness.
The results of a comprehensive program of tests carried out to investigate
the influence of particle impact angle are presented in Figure 21.13 [3]. 0-2 in alu-
minum oxide particles were impacted against a steel target, which is the reference
point in this particular program of work, and so the data for 90° impact is the same
as that presented earlier in Figures 21.7, 21.9, 21.11 and 21.12. It will be seen
from Figure 21.13 that there is little change in the response to degradation until the
impact angle is below about 50°. There is than a very marked difference in per-
formance with only small incremental changes in impact angle.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

616 Chapter 21

100
t/D

—
'4 80
C3
n_
jj 60
o Particle Size - 0-2 in
| 40Target Material - Steel
'o
« 20
£

2000 3000 4000 5000 6000

Particle Velocity - ft/min

Figure 21.13 The influence of particle velocity and impact angle on the degradation of
aluminum oxide particles

With a decrease in particle impact angle it would appear that there is little
change in the particle velocity at which the onset of degradation occurs. The tran-
sition from zero degradation to total degradation, however, becomes an increas-
ingly more gradual process as the particle impact angle reduces. At impact angles
of 15° and 20° it would appear that this transitional process will be spread over a
very wide range of velocity values. At an impact angle of 10°, however, there is a
significant change once again, in that no particle degradation was recorded at all
up to 6000 ft/min. In Figure 21.14 an alternative plot of the data from this program
of tests is presented.

Particle size - 0-2 in

Particle velocity - 4600 ft/min
Target material - Steel

10 20 30 40 50 60 80 90
Particle Impact Angle - a - degrees

Figure 21.14 The influence of particle impact angle on the degradation of aluminum
oxide particles.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 617

This is effectively a slice taken from Figure 21.13 at a particle velocity of

4600 ft/min. It will be seen from this that tests were carried out at regular incre-
ments of impact angle of about 10° between 10° and 90°. This plot shows quite
clearly that at impact angles below about 12° no degradation occurs, and that at
impact angles above about 55° the degradation remains essentially constant at the
maximum value for this particular impact velocity.

2.6 Other Variables

Segler [4] investigated the influence of moisture content on particle degradation
and showed that degradation can increase dramatically with decrease in moisture
content. The results of the following three tests with peas show the sensitivity to
this variable:

Moisture Content-% 17-1 16-1 15-4

Broken Particles -% 0-1 1-1 11-1

Segler investigated the effect of particle concentration and found that the
damage decreased as the solids loading increased. The damage produced when the
peas were introduced individually was four times higher than in dense flow. A
similar effect is found in erosive wear and can be attributed to the cushioning ef-
fect of dense flows.
He also examined the damage to the peas in identical pipelines having
nominal bores of 2 and 10 inch. It was found that the damage in the 2 in bore pipe-
line was two to three times greater than that in the 10 in bore pipeline. His expla-
nation was that the frequency of pipe wall impacts, for such large particles, would
be more frequent for the small bore pipeline.

3 RECOMMENDATIONS AND PRACTICAL ISSUES

The results from the various programs of work reported here have produced some
very interesting relationships with respect to many of the variables investigated,
and should provide useful guidance to the design engineer who has to ensure that
material degradation is reduced to a minimum in pneumatic conveying system
pipelines.

3.1 Particle Velocity

Particle velocity has been a major consideration in this work and it has been
shown quite clearly that there is a threshold value of particle velocity below which
no degradation occurs. The value of this particle velocity for the aluminum oxide
was about 2000 ft/min and was influenced only slightly by particle size, target
material, and particle impact angle above about 15°.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

618 Chapter 21

3.1.1 Dense Phase Conveying

At velocities only slightly lower than this, however, the mode of conveying
changes from dilute phase, suspension flow, to dense phase, non-suspension flow,
for many of those materials capable of being conveyed in dense phase. In dense
phase conveying little impact occurs in horizontal pipelines and the mode of con-
veying mostly involves sliding of the particles through the pipeline. With materials
having good permeability, conveying is in plugs and slugs, and for materials hav-
ing good air retention, it is as a moving bed along the bottom of the pipeline.
When particles slide through the pipeline the interaction results in attrition
rather than degradation. In dilute phase there may be little particle to pipe wall
interaction, and it is suspected that most of the damage results from impact with
pipeline bends. In dense phase, although the velocity is low, there is a significant
amount of particle to pipe wall interaction and for certain materials this is likely to
cause more damage to the particles than the bends.
As a consequence it is possible for some materials to suffer a greater amount
of damage in low velocity dense phase flow than they would in higher velocity
dilute phase flow. It is important, therefore, to examine the relative effects of deg-
radation and attrition on the conveyed material before deciding upon the type of
pneumatic conveying system to be employed.

3.1.2 Dilute Phase Conveying

For many materials dense phase conveying is not an option, for the majority of
materials can not be conveyed at low velocity in a conventional conveying system.
For these materials conveying has to be in suspension flow and so if the material is
friable, degradation must be limited. To this end the material should be conveyed
at as low a velocity as possible, consistent with reliable conveying, and the pipe-
line should be stepped to a larger bore part way along its length to reduce the high
conveying air velocities that result at the end of the pipeline.
With a 15 lbf/in2 pressure drop in a positive pressure system, discharging to
atmospheric pressure, the conveying air velocity will approximately double from
the material feed point to discharge. For the situation presented in Figure 21.7 it
will be seen that at 2000 ft/min no damage occurs, but at 4000 ft/min 80% of the
particles are broken.
As the air expands through the pipeline, therefore, it is the bends at the end
of the pipeline, in a single bore line, that are likely to cause the majority of the
damage. By stepping the pipeline the maximum velocity in the pipeline could pos-
sibly be limited to 3000 or 3200 ft/min, at which the degradation would be limited
to only 30%.

3.2 Particle Impact Angle

For given conveying conditions, particle impact angle is probably the most impor-
tant variable with respect to pneumatic conveying system pipelines. The angle of
impact of particles against pipeline walls will generally be very low since particles

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 619

will only suffer a glancing impact. From the data presented here it would appear
that little degradation will occur in straight pipeline, even for long pipelines and
repeated impacts.
It is clearly major changes in flow direction, and in particular bends, that are
likely to result in the majority of degradation occurring. In this respect, particle
impact angle can be related approximately to the radius of curvature of a bend. In
a short radius bend the particles will impact at a high value of angle, but in a long
radius bend the impact angle will be much lower, as illustrated in the previous
chapter with Figure 20.14. Since degradation reduces significantly with reduction
in impact angle, the use of long radius bends would be recommended in any sys-
tem where particle degradation needs to be minimized.

3.3 Bend Material

The choice of material for the pipeline, and in particular the bends, provides an-
other means by which particle degradation can be minimized. Although there is
little change in the value of the lower threshold velocity, below which no degrada-
tion occurs, with respect to target material, there is a very significant effect on the
upper threshold value.
Thus, for a given particle impact velocity, very much less damage will result
to particles if they impact against a surface such as Plexiglas or aluminum, than
will occur if they impact against steel or glass. If it is possible to use a more resil-
ient material, such as rubber or polyurethane, an even more significant reduction
in particle degradation may be achieved.

4 PNEUMATIC CONVEYING DATA

To provide some data on the potential order of magnitude of the problem of deg-
radation, for materials conveyed in dilute phase suspension flow in a pneumatic
conveying system, four different materials were pneumatically conveyed and the
resulting degradation was monitored. A large scale pneumatic conveying facility
was used and on-line samples were taken for analysis. Each material was re-
circulated through the test loop a number of times so that the influence of convey-
ing distance could also be investigated [9],

4.1 Material Degradation Data

The data has, in fact, been presented earlier, in the chapters dealing with convey-
ing data on specific materials. Thus data on coal will be found in Chapter 10 at
section 5 and with Figure 10.28. Data on both sodium chloride (common salt) and
sodium carbonate (soda ash) were included in Chapter 11 around Figures 11.5 to
11.7. The fourth material was silica sand and this was included in Chapter 14 with
Figures 14.6 and 14.7.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

620 Chapter 21

4.2 Conveying Characteristics

The influence that pneumatic conveying can have on the conveying characteristics
of a material has been considered for many different materials. As a result of re-
circulating materials, in order to derive conveying characteristics, some of the
more friable materials have degraded to such an extent that the conveying charac-
teristics for the material have changed significantly. It must be recognized that the
pneumatic conveying of a friable material can also change the conveying charac-
teristics of a material.

5 PARTICLE MELTING

Particle melting is a form of material degradation that often occurs in pneumatic

conveying plant handling plastic type materials, particularly in pelletized form. If
conventional pipeline is used, materials such as polyethylene, nylon and polyesters
can form cobweb-like agglomerates. They are variously given names such as 'an-
gel hairs', 'raffia', 'snake skins' and 'streamers'.
They frequently cause blockages at line diverters and filters which require
plant interruption to remove them. Equipment is generally installed at the termi-
nating end of the system for this purpose. Such equipment is necessary because
they also cause material rejection by customers, for the presence of these contami-
nants in the product is undesirable.

5.1 Mechanics of the Process

The streamers are caused by the pellets impacting against the bends and pipe
walls. A considerable amount of energy is converted into heat by the friction of the
two surfaces when they touch. If the surface of the pipe is smooth, the pellet will
slide. This contact, though momentary, decelerates the particle by friction which is
transformed into heat. This is generally sufficient to raise the temperature at the
surface of the pellet to its melting point. To a certain extent this is analogous to the
thermal model proposed for erosive wear.

5.2 Influence of Variables

The onset of the formation of these angel hairs or streamers is the result of a com-
bination of conditions. Particle velocity is the most important, but it also depends
upon the temperature of both the pipe and the pellets, and the solids loading ratio
of the conveyed material.
The influence of conveying line exit air velocity for low density polyethyl-
ene is shown in Figure 21.15 and the influence of solids loading ratio for this same
material is given in Figure 21.16 [10]. In each case the degradation of the material
is expressed in terms of the mass of streamers and fines produced, in ounces, per
ton of low density polyethylene conveyed.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

 Particle Degradation 621

Material:
Low Density Polyethylene
Pipeline:
Bore - 4 in
Material - Aluminum
Surface - Sandblasted

Conveying Conditions:
Solids Loading Ratio = 8
Material Temperature = 120°F

0
4000 6000 8000 10,000 12,000
Conveying Line Exit Air Velocity - ft/min

Figure 21.15 Influence of velocity on the degradation of LDPE.

5.3 Pipeline Treatment

The formation of streamers and fines can be reduced quite considerably by suita-
bly treating the pipe wall surface. A roughened surface is necessary in order to
prevent the pellets from sliding. If the surface is too rough, however, small pieces
will be torn away from the pellets instead, and a large percentage of fines will re-
sult. It will also have an adverse effect on the pressure drop, and hence on material
conveying capacity.

Conveying Conditions:
Exit Air Velocity = 8000 fl/min
o Material Temperature = 120°F
o

Material:
Low Density Polyethylene
O
Pipeline:
Bore - 4 in
Material - Aluminum
Surface - Sandblasted

4 6 10
Solids Loading Ratio

Figure 21.16 Influence of solids loading ratio on the degradation of LDPE.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.

622 Chapter 21

Although the results presented in Figures 21.15 and 16 were obtained from
tests carried out with pipe surfaces roughened by sand blasting, this treatment is
not recommended as it will result in the generation of a large percentage of fines.
Also, this roughness is relatively shallow in depth and an aluminum surface will
wear so that the pipe must be retreated in six to twelve months [10].

REFERENCES

1. D. Mills and J.S. Mason. Problems of particle degradation in pneumatic conveying

systems. Proc Pneumotransport 4. Paper F3. 10 pp. BHRA Conf. Carmel, California.
June 1978.
2. D. Mills and J.S. Mason. The effect of pipe bends and conveying length upon particle
degradation in pneumatic conveying systems. Proc 3 rd Powder and Bulk Solids Conf.
Chicago. May 1978.
3. A.D. Salman, M. Szabo, I. Angyal, A. Verba, and D. Mills. The design of pneumatic
conveying systems to minimize product degradation. Proc 13th Powder and Bulk Sol-
ids Conf. Chicago. May 1988.
4. G. Segler. Pneumatic grain conveying with special reference to agricultural applica-
tions. Germany. 1951.
5. G.P. Tilly. Erosion caused by solid particles. Treatise on Materials Science and Tech-
nology. Vol 13, pp 287-319. Academic Press Inc. 1979.
6. G.P. Tilly and W. Sage. The interaction of particle and material behavior in erosion
processes. Wear, Vol 16, pp 447-465. 1970.
7. A.D. Salman, A. Verba, and D. Mills. Particle degradation in dilute phase pneumatic
conveying systems. Proc 17th Powder and Bulk Solids Conf. Chicago. May 1992.
8. D. Mills. Particle degradation in pneumatic conveying, pp 31-48. Freight Pipelines. Ed
G.F. Round. Elsevier. 1993.
9. D. Mills. The degradation of bulk solids by pneumatic conveying and its simulation
by small scale rigs. BMHB report. Feb 1988.
10. J. Paulson. Effective means for reducing formation of fines and streamers. Proc Conf
on Polyolefins. Soc of Plastic Engineers. Houston. 1978.

Copyright  2004 by Marcel Dekker, Inc. All Rights Reserved.