Uploaded by rafaeldalibera

Characterization of rotational molding grade polyethylene powders

Characterization of
Rotational Molding
Grade Polyethylene
Powders"
J . L. Throne and M . 4 . Sohn
Department of Polymer Engineering
University of Akron
Akron, Ohio 44325
ABSTRACT
Rotational moMing grade polyethylene powders are
classijied according to sieve size, to typically -35 mesh,
and by flowability, usually determined by pouring freeflowing powder from a funnel onto aJixed base cone.
In the latter case, the static angle of repose is then
measured. This study concentrates on general methods
of characterizing powders and relating these methods
to flowability of five selected commercial rotational
molding grade polyethylene powders. The techniques
described in this study include methods for measuring
particle size distribution,particle shape characteristics,
static and dynamic angles of repose, powder shear
strength, and angle of internal friction. All powders
examined had similar particle size distribution profiles,
with distributions skewed to the coarser, -35 mesh size
particles. The stutic and dynamic angles of repose were
quite similar and the latter were essentially independent
of the volumefraction ofpowder in the rotating cylinder
and of the speed of rotation of the cylinder. Allpowders
had shape parameters-elongahonal ratio, chunkiness,
circularity, and roundness-that were essentially similar and independent of particle size, and the compressibilities of the bulk powders were not strongly dependent
on powder bed height.
INTRODUCTION
where a combination of water spray and ambient air is
directed against the mold surfaces, cooling them and
hence the polymer adhering to the inner mold walls. The
cycle ends when the cooled parts are removed from the
molds.
Nearly all rotational molding applications use polyethylenes. Despite the mature nature of the process, there
are relatively few guidelines for determining when a resin
is suitable for rotational molding. It is known that powders with particle sizes less than 200 mesh (sieve screen
size or about 50 pm in dimension) tend to fluidize excessively and to bridge in acute-angled comers. It is
known that powders with particle sizes greater than about
The rotational molding process is shown schematically
in Figure 1. Coarse polymer powder is measured into
open clam-shell molds. The molds are mechanically
clamped shut and rotated about their equatorial and polar
axes. The rotating mold frame is then inserted into an
isothermal forced air convection oven and held there until
the polymer melts and coats the inner surface of the mold
walls. The mold frame is then moved to a cooling area
'Mr. M.-S. Sohn received a research assistantship from Association of Rotational Molders through the Rotational Molding Development Center. University
of Akron, Dr. James L. White, Center Director.
Advances in Polymer Technology, Vol. 9, No. 3, 181-192 (1989)
0 1989 by John Wiley & Sons, Inc.
CCC 0730-6679/89/030181-12$04.00
POLYETHYLENE POWDER CHARACTER
FIGURE 1 Schematic of rotational molding machine.
35 mesh (sieve screen size or about 500 pm in dimension)
require long oven times and finished parts tend to have
greater porosity than finer powders. From a cursory examination of powders that rotationally mold well and
poorly, it was concluded that powder shape was also
important. Spherical and accicular or fiberlike particles
did not densify as readily as powders having particles of
near-cubic or “squared-egg” shapes. Easy-flow powders
yield more uniformly walled parts than do powders that
are sticky, tacky, or tend to bridge.
Traditionally, the rotational molding industry relies on
powder sieve size and flowability to characterize powders. The latter is frequently related to the static angle
of repose of freely poured powder. It does not appear
that characterization of rotational molding grade polymer
powders has been extended beyond these guidelines. This
paper characterizes particle sizes and shapes for five commerical rotational molding grade polyethylene powders.
’
Materials
Five polyethylene polymer powders listed in Table I
have been examined in detail. All of these are commercial
grade polymers.
Particle Size and Shape Characterization
The size and shape characteristics of the resin powders
were determined in several ways. Table I1 gives a sum182
mary of the methods used. Note that not all resins were
evaluated in all tests. In several cases, the experiments
were done for a fee at outside laboratories. In other cases,
the experiments were quite involved. Time and money
allowed only one or two resins to be characterized in
these cases.
The techniques described below include measuring
techniques for static and dynamics angles of repose, analytical techniques for determining particle sizes, size
distributions, and shapes, determination of powder internal angle of friction and shear characteristics, and bulk
compressible characteristics. These techniques are described in detail in Ref. 2.
Size Distribution
Four techniques were used to determine powder size
distribution. These are listed in Table 111. Mechanical or
physical shaking of the powder through sieves of graduate opening sizes (mesh sizes) is a common method of
classifying rotational molding grade powders. The five
resin powder size distributions are shown in Figure 2.
All resins show distributions that are skewed toward - 35
mesh. The Gilson sonic sifter uses a pulsed column of
air to fluidize the powder particles, which then settle
through screens of graduated mesh size. An ancillary
mechanical pulse or tap is also used. The cumulative
particle size distributions for two of the commercial resin
powders are shown in Figure 3(a).
The Leeds and Northrup Microtrac (tm) uses light scatVOL. 9, NO. 3
POLYETHYLENE POWDER CHARACTER
TABLE I
Rotational Molding Grade Polyethylenes Used in This Study"
ROpertY
Unit
Density
Tensile strength
Impact ( - 40°C)
Melting point
Ultimate elongation
g/cm3
psi
ft-lb
(thickness)
"C
SH
SL
NH
NL
0.952
2360
395 (1 in.)
0.939
3900
440 (1 in.)
0.95
3300
48 (1/8 in.)
0.94
2560
45 (118 in.)
-
-
-
-
126.5
136
600
%
765
UL
0.939
2500
46.5 (118 in.)
-
-
~~~
%H = Esso HDPE (HD-8761 Escorene); SL = Esso LLDPE (LL-8461 Escorene); NH
LLDPE (MN 720 Microthene).
TABLE I1
Methods Used to Characterize Rotational Molding
Grade Polyethylene Powders
0
0
0
0
0
Size distribution
Shape characteristics
Angle of repose
Shear properties (angle of internal friction)
Bulk density measurement
=
Enron HDPE (5305T);NL = Enron LLDPE (5304T);UL .= US1
The four techniques are compared in Figure 3(d) for
one material, Esso HDPE. As is apparent, these techniques do not yield identical results. For example, the
50% CP value for the Gilson unit is at less than 200 pm
whereas for the Microtrac, it is about 270 pm.
Particle Shape Analysis
tering to measure particle size. The laser light scattered
by particles flowing through an orifice is passed through
a set of filters, as shown in schematic in Figure 4. The
cumulative size distributions of four powders, each scanned
three times, are shown in Figure 3(b). The cumulative
percentage curves for the two Esso resins in Figures 3(a)
and 3(b) compare well. The Elzone (tm) particle size
analyzer uses electrical detection of individual particles.
Individual particles passing through an orifice in an electrical field cause generation of electrical pulses. The amplitudes of these pulses are proportional to the volumes
of the particles. Fluid flow rate, orifice size, and particle
concentration are parameters. The Elzone-generated cumulative percentage curve for Esso HDPE in Figure 3(c)
shows a 50% cumulative percentage (CP) values of less
than 200 Fm. This agrees better with the Gilson data
than with the Microtrac (tm) data.
TABLE I11
Techniques Used to Determine Rotational Molding
Grade Polyethylene Powder Size Distribution
Methods
Principles
Physical sieving
Microtrac
Elzone
Gilsonic sifter
Mechanical shaking of sieves
Light scattering from particles
Detection electrical pulse of particle
Vertical oscillating column of air
ADVANCES IN POLYMER TECHNOLOGY
As noted, the recommended rotational molding powder particle has a near-cubic shape. The literature on
particle shape analysis has grown rapidly recently with
the development of computer-aided scanning devices.
Several sets of nomenclature are currently used to describe particle shape. One of the oldest is the Heywood
criteria: where the dimensions of an irregular particle
are referred to a projected area diameter and sieve aperture. Details of the Heywood criteria are given in Table
IV. The primary shape factors are elongational ratio,
flatness ratio, sphericity and roundness. Recently, these
shape factors have been augmented and superseded by
other sets of terms. Beddow and Phillip4 established criteria based on describing the boundary of the two-dimensional projection of the particle in terms of a Fourier
series of the radius and subtended angle, R(8). These
descriptors are then used to determine secondary shape
factors such as first and second moments of radial distribution means, and a perimeter descriptor. Fong et al.5*6
noted that the Beddow shape factor criteria were not
invariant to particle orientation. They suggested a new
set of invariant morphic descriptors, based on Fourier
series expansion of normalized dimensions. These allow
for accurate description of reentrant particles. Mendelbrot'
considered the shape-filling nature of a highly convoluted
or rugged curve to be described by a mathematical dimension that is not integral. This is called the fractal
dimension. Also, Kayes proposes adding two additional
183
POLYETHYLENE POWDER CHARACTER
100,
1 4 0 1
5
0
F, ,
,
~
~
20
020
50
80
I110
I
1
140
20
so
80
Sire (mesh)
110
140
Size (mesh)
*
o
o
l
60
a
s
0
20
I
20
50
I30
110
140
Size (mesh)
Sire (mesh)
100
1
Size (mesh)
FIGURE 2 Mechanical sieve size distributions for rotational molding grade polyethylene powders.
dimensional ratios, external compactness and chunkiness, to the three Heywood criteria. These are defined
in Table IV.
In this study, the sieve size-fractionated powder particles were sparsely sprinkled onto an epoxy-coated metal
stand. The particles were then gold-sputtered and scanning electron microphotographed.These photographs were
then optically raster-scanned and the shape data analyzed
184
using Image-Pro (tm), a shape analyzing software package.9Four particle shape factors-elongational ratio, circularity or projected sphericity, chunkiness, and roundness factor-are given in Figure 5 as functions of the
sieve mesh size for all five powder types. All powders
show similar shape characteristics. More importantly, the
shape characteristics appear to be relatively independent
of particle size.
VOL. 9, NO. 3
POLYETHYLENE POWDER CHARACTER
100
bQ
u
m
80
0
"
C
2 60
a
0)
.?
"
-
40
0
0-0
3
5
0
ESSO HOPE
ESSO LLDPE
.....
20
0 ..
0
0
100
200
(a)
300
400
500
Size (micron)
I
100
n
0
Cn
80
"
C
u
v
L
60
u
Q
0
Materiels
2
40
&
-03
5
0
20
/
0-0
ESSOHDPE
0. . . 0 ESSO UDPE
/8'
aF
ENRON HDPE
US1 U P €
A-A
A--A
0
100
200
300
400
500
600
700
Size (micron)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .._
...................................
-i..............................
' /
:!:!:!--............................
I
.
.
I
I
-
.......................
.. _. . . . . . . . . . . . . . . . . . . . . .
... .,, ... ._. . . . . . . . . . . . . . . . . . .
, -
......................
.:-
....
.-:
-
... .:.,. ,... :.... .... :.... _... . . . . . . . . . . . . . . ... .: . . .. .. . .. .. . . . . . . . . . . . . . .
. .! .. :1 :. ; : :. j: T
.. . .. .. . . . . . . . . . . . . . . . . .-:
I
.....
,
.
, . . : ' i: :. i :: .T : .
:;:::/:j :i:
I
...........i
. . . . . . . . . . .-::=
.. !.,. . .! .... .. .... ! ..,../ .. .T i . . . . . . . . . . i
. .. . ... ... .. ... ..... ... ..... ... _. . . .. . . ..__
I
.
A
(4
!
.
I
A
Size (micron)
FIGURE 3 Cumulative percentage distributions using several techniques for rotational molding grade polyethylene powders. (a) Gilson
and Esso LLDPE (0).(b) Microtrac (tm) technique for Esso HDPE (0),Esso LLDPE (O),
sonic sifter technique for Esso HDPE (0)
Enron HDPE (A), and US1 LLDPE (A).
(c) Elzone (tm) technique for Esso HDPE.
ADVANCES IN POLYMER TECHNOLOGY
185
POLYETHYLENE POWDER CHARACTER
FIGURE 3 (d) Comparison of cumulative percentage distributions for Esso HDPE for mechanical sieving (l), Gilson Sonic Sifter (2),
Microtrac (tm) (3),and Elzone (tm) (4).
Flowability
TABLE IV
Particle Shape Characteristics
Flowability of rotational molding powders is determined primarily from the static angle of repose. This is
done by pouring dry powder through a funnel to a fixed
cone height [Figure 6(a)] or to a fixed cone base dimension [Figure 6(b)]. Powder forms a cone having an equilibrium angle, called the static angle of repose. The-tilting
table [Figure 6(c)] is sometimes used but this test tends
to give false values for powders that interlock or bulk
readily. The static angles of repose of the five powders
are given in Table V. In rotational molding, the powder
is usually in a dynamic, flowing condition. The dynamic
angle of repose is defined as the equilibrium angle the
length
Elongational ratio = breadth
Chunkiness = reciprocal of elongational ratio
radius of sharpest comer
Roundness Factor =
longest comer
Circularity
- circumference of circle with same projected area
actual circumference of projected area
External compactness
- square of diameter of equal area to that of profile
square of diameter of embracing circle
Rolaling Filler
Slalionary Filler
Sample Cell
Pholodeleclor
Lasei‘Beam
Lens
AizilT
Microcomputer
Printer
FIGURE 4 Schematic of microtrac (tm) particle size characterization.
186
VOL. 9, NO. 3
2
.0
c
x
'i:
i;j
2
0
01
0
.2
-0
0
.c
10
0.68
0.72
0.76
0.80
1 .o
1.5
2.0
2.5
0
0
A
fi
A
B
40
0
O
A
0
A
70
m
0
A
0
A
A
0
0
A
0
0 ESSO HDPE
0 ESSO LLDPE
A ENRONHDPE
100
A ENRON LLDPE
0 US1 LLDPE
0
n
A
0 ESSOHDPE
0 ESSO LLDPE
A ENRON HDPE
A
0
Size (mesh)
S
0
0
A
0
130
O
[t:
3
8
z
0
.c
7
O.lO+
0.201
0.30-
0.40
10
0.2 J
0.4--
0.8
0
A
0
0
40
A
m
70
100
A ENRON LLDPE
0 US1 LLDPE
0 ESSO HDPE
0 ESSO LLDPE
A ENRON HDPE
A
0
Q
US1 LLDPE
A ENRON LLDPE
0 ESSO HDPE
0 ESSO LLDPE
A ENRON HDPE
Size (mesh)
a
m
1
POLYETHYLENE POWDER CHARACTER
V
V
FIGURE 6 Methods of determining angles of repose: (a) fixed height cone, freely poured powder; (b) fixed base cone, freely poured
powder; (c) tilting table; (d) rotating cylinder, dynamic angle of repose.
powder forms while in a steady flowing condition. The
dynamic angle of repose can be measured by rotating the
powder in a horizontal cylinder [Figure 6(d)]. This was
done for all five powders, with the relative powder volume and the cylinder speed of rotation as parameters.
An example of the technique is shown in Figure 7 for
Esso HDPE at 4 rpm at 5% powder volume. It was found
that the dynamic angle of repose for each of the five
powders was essentially independent of volume in the
range of 5-30% and essentially independent of cylinder
speed of rotation in the range of 4-1 1 rpm. The average
dynamic angles of repose of the five powders are given
in Table 111. The values average around 40". This is
consistent with other polymer powders and supports the
similarities in the particle shape, as seen in Figure 5.
Powder Internal Friction Angle
Rauwendaal" classifies polymer powders as either
Coulomb flow powders or viscous flow powders. In the
former, the particles remain in continuous contact with
their neighbors in any flow situation. In the latter, internal
forces are resisted by momentum transfer between particles that move relative to one another. Kurihara" proposes that the flowability of powder in rotating systems
can be characterized in terms of a viscosity. In his analysis, he considers the powder bed to be a steady state
Newtonian fluid. He derives a relationship between the
angle of inclination of the free surface of the powder
"fluid," the rate of rotation and the volume of powder in
the container. From this, he is able to obtain a fluidlike
FIGURE 7 Example of measuring dynamic angle of repose for Esso HDPE, at 5% (vol), and 4 rpm.
188
VOL. 9, NO. 3
POLYETHYLENE POWDER CHARACTER
TABLE V
Static and Dynamic Angles of Repose and Angles of Internal Friction for
Rotational Molding Grade Polyethylene Powders
Average dynamic angle of repose
sd"
Average static angle of repose
sd"
Angle of internal friction of PE powder by shear cell
ES HDPE
ES LLDPE
EN HDPE
EN LLDPE
US1 LLDPE
41.3
1.82
40
1.17
30.3
39.6
I .so
40
1.79
31.8
41.4
2.63
40
1.47
32.9
44.4
4.11
42.3
2.49
40
2.19
34.0
45
2.48
35.3
'Standard deviation.
viscosity for powder. The Jenike shear cell is used to
determine the shear strength of powders under varying
normal loads. As seen in Figure 8, a strain gage is used
to measure the force required to shear the powder contained in the cell at a straining rate of lO%/min. The
shear stress as a function of horizontal displacement, with
applied normal force as parameter, is shown for each of
the five powders in Figure 9. Note that, in all instances,
the shear stresses plateau. These maximum values are
replotted in Figure 10 for each of the five powders. Note
that, for every powder, the limiting shear strength of the
powder, when there is no applied normal load, is zero.
This means that these rotational molding grade powders
can be considered as cohesionlesspowders. Furthermore,
the slopes of these shear strength curves are the tangent
values of the angles of internal friction. These values are
given in Table V. It is again apparent that all of these
powders have similar characteristics, being cohesionless
and with internal frictional angles between 30" and 34".
Bulk Density
In the rotational molding process, the recirculating
powder is alternately fluidized and compacted. Very fluffy
powder can become a problem in processing. A simple
experiment to determine height-dependent tamped and
untamped bulk density was devised. The volume of particles in a graduated cylinder is measured each time 2 g
of powder is added to the particles in the cylinder. If the
powder is to be tamped, it is tamped between each powder addition. The effect of powder height on the tamped
and untamped bulk density of Esso HDPE powder is
shown in Figure 11. It is apparent that this powder does
not show appreciable compaction due to powder height.
A simple least-squares fit to the data shows that the following equation fits both tamped and untamped conditions:
Y
=
a - b * exp(-c. X)
FIGURE 8 Schematic of Jenike shear strength measuring device.
ADVANCES IN POLYMER TECHNOLOGY
189
POLYETHYLENE POWDER CHARACTER
N
15
.-C
1
P
10
n
n
u
-
u)
L
B
m
r
O i
0.0
0.1
0.2
0.3
0.4
Horizontol Di~placcmsnt(in)
5
I
04
0.5
0.0
I
0.2
0.1
0.3
0.4
0.5
Horizontal Diaplocsmmt (in)
20
~0.0
0.1
0.2
0.4
0.3
0.5
Horizontal Diapbcsmsnt (In)
20 7
I
I
0.0
0.1
0.2
0.3
0.4
0.5
Horizontol Di*placement (in)
FIGURE 9 Displacement-dependentshear stress for rotational molding grade polyethylene powders. Applied normal mass N (kg):
(0)
16; (0)32; (A)48;
64; (a) Esso HDPE; (b) Esso LLDPE; (c) Enron HDPE; (d) Enron LLDPE; (e) US1 LLDPE.
(A)
where y is the powder density, x is the powder height,
and a, b, and c are curve-fit constants. This equation is
well known in powder processing."
CONCLUSIONS
This work characterizes particle sizes and shapes for
five commerical rotational molding grade polyethylene
powders and concludes that:
190
All powders show skewed particle size distribution
toward the -35 mesh.
Microtrac (tm) and mechanical sieving yield similar
particle size distribution curves, whereas Gilson sonic
sifter and Elzone (tm) data show similar particle size
distribution curves. The 50% cumulative percentage
values for Esso HDPE powder are similar for these
particle size measuring methods, but are not identical.
From examination of powder particle shape characteristics, it was determined that the four primary shape
VOL. 9, NO. 3
POLYETHYLENE POWDER CHARACTER
"n
G
I
18.00..
ca
cn. 18.001
18.OOt
Y
Y
s
8
01
01
E
z
5
iz.oo--
12.00-
VI
L
L
0
0.00
0
8
16
24
32
24
32
c/
0
Normal Load (I.)
0
16
Normal Load ( X I )
24
24.00
0
0
I6
Normol Load ( 11)
Normal Load ( ce )
24.00
Normal Load (Ki )
FIGURE 10 Load-dependent shear strength for rotational molding grade polyethylene powders; (a) Esso HDPE; (b) Esso LLDPE;
(c) Enron HDPE; (d) Enron LLDPE; (e) US1 LLDPE.
0
characteristics-elongational ratio, chunkiness, circularity, and roundness factor-were essentially independent of particle size.
Further, it was determined that all five powders had
quite similar particle shape characteristics. That is,
there was no apparent evidence of accicular or reentrant particles.
From free-pouring and cylinder rotating experiments,
it was determined that the ranges in values for static
ADVANCES IN POLYMER TECHNOLOGY
0
0
0
and dynamic angles of repose of these powders were
40-45".
From shear box experiments, it was determined that
the range in values for angles of internal friction of
these powders was 30-35".
Also from shear box experiments, it was determined
that all these powders are cohesionless.
It was found that for one rotational molding grade
powder, there was relatively little compaction due to
191
POLYETHYLENE POWDER CHARACTER
./.--.-
CI
n
m
0
n
~-o/o-o-o)
0-0
.
m-m
U n l a m p i n g Condition
T a m p i n g Condition
m
0
0.2
3
6
2
9
12
REFERENCES
1. M. A. Rao and 1. L. Throne, Polym. Eng. Sci., 12, 237 (1972).
2. M . S . Sohn, Master of Science Thesis, Department of Polymer Engineering,
University of Akron, Akron OH 44325, 1989.
3. H. Heywood, J . Pharm. Pharmacol., 16, 50 (1963).
4. I. K. Beddow and G. C. Phillip, Planseeberi. Pulvermerall. 23, 3 (1975).
5. S.-T.Fong, 1. K. Beddow, and A. F. Vetter, Powder Techno/., 22, 17
( 1979).
6. 1. K. Beddow, Planseeberi. Pulvermerall. 21, 3 (1979).
7. B. P. Mendelbrot, Fractals. Form. Chance and Dimension, Freeman, New
York, 1979.
8. B. H. Kaye, Proceedings PowderMerallurgy Conf., Washington, DC,June
1980.
9. Image-Prof, Microcomputer Software Package, Media Cybernetics, Inc.,
4
15
Height (cm)
FIGURE 11 The effect of powder bed height on bulk density
of Esso HDPE rotational molding grade polyethylene powder:
(0)untamped condition; (0)tamped condition.
192
powder height for both tamped and untamped bulk
powder
Silver Spring, MD.
10. C. Rauwendaal, Polymer Extrusion. Hanser, Munich, 1986.
1 1. K. Kurihara, Oyobutsuri, 34, 727 (1965).
12. J. R. Johanson and A. W. Jenike, Bulletin 116, Utah Engineering Experimental Station, Bull. Univ. Utah, 53, 21 (1962).
VOL. 9, NO. 3