Improvement of Processing Parameters by Using

Processing Aids During the Linear Low-Density

Polyethylene Film Blowing Process

Abbas Fatehi, Naser Sharifi Sanjani

School of Chemistry, Faculty of Science, Tehran University, Tehran, Iran

Polymer processing aids are used to improve process- but when it contains HFP units, it begins to be disordered
ing properties in the polyethylene industry. These mate- and becomes a material similar to a noncrosslinked elasto-
rials improve not only the physical and mechanical mer. Melting temperature will, therefore, be increased if
properties of the final products but also their process-
ing properties. This paper studies some of the process- the concentration of HFP in the copolymer is increased
ing variables such as die pressure, melt temperature, [3].
masterbatch activity, and die gap by examining the The PPAs can affect mechanical and optical properties
functions of polymer processing aids and, last but not by eliminating sharkskin, improving surface smoothness,
least, the effects on the film blowing process of two- improving impact resistance, reducing operation pressure
component processing aids containing a perfluorinated
additive and polyoxyethylene. J. VINYL ADDIT. TECHNOL., and amperage, and increasing output in film blowing
16:78–83, 2010. ª 2009 Society of Plastics Engineers extrusion. In film blowing extrusion the concentration of
processing aid is 3–5%.
Other materials are applied as processing aids, exam-
INTRODUCTION ples of which are polyethylene, waxes, boron nitride, etc.
Linear low-density polyethylene is a random copoly- Some of them are also lubricants. Lubricants can be clas-
mer of ethylene and an alpha-olefin monomer. Alpha-ole- sified into two categories, internal and external [4]. Exter-
fins enter the polyethylene structure as side branches. The nal lubricants are incompatible with the polymer and
structure of linear low-density polyethylene, therefore, migrate to inter-border surfaces, but they do not adhere to
contains linear chains as well as many short side branches the extruder metallic surface. External lubricants can
[1]. Furthermore, its molecular-weight distribution is nar- decrease apparent viscosity by reducing friction, and they
rower. So, the sensitivity of linear low-density polyethyl- show a few of the processing aid properties. Poly(ethylene
ene viscosity to shear rate will be less, and its processing glycol) (PEG) is used as an external lubricant. It is in-
will face some problems. At higher shear rates, the poly- compatible with the polyolefin because of its functional
mer exhibits melt instability. This instability is classified group and thus migrates to the inter-border surface. How-
into two types. The first instability develops in the en- ever, because of the lower electronegativity of oxygen as
trance of the die as melt fracture. The second one devel- compared to fluorine, it cannot react to form metallic
ops at high shear rates as exhibited in the exit gap of the oxides and thus exits from the extruder together with the
die. In this kind of instability, the diameter of the output melted polymer [5].
material exuding from the die is higher. An example of Taguchi experimental design differentiates between
these kinds of instabilities is sharkskin [2]. Perfluorinated control factors and noise or uncontrollable factors and
polymers are a group of polymer processing aids that treats them separately by means of special design matri-
were discovered by the DuPont company in the 1960s. ces called ‘‘Orthogonal Arrays’’ (OAs). Columns and
The perfluorinated processing aids (PPAs) used are rows of an OA are arranged in a fixed way indicating
copolymers of vinylidene fluoride with hexafluoropropy- the combination of factor levels of each experiment to
lene (PVDF-HFP). The mole fraction of fluorine in this be run and allowing the simultaneous evaluation of
copolymer is 0.615. The PVDF has a crystalline structure, several parameters with the minimum number of trials.
The main advantage of the Taguchi parameter design, as
opposed to the classic factorial design methods, lies in
Correspondence to: A. Fatehi; e-mail: afatehi@khayam.ut.ac.ir
the introduction of noise factors into the experimentation
DOI 10.1002/vnl.20210
Published online in Wiley InterScience (www.interscience.wiley.
which provoke an uncontrolled variation leading to
com). a noise incentive response and therefore to higher

2009 Society of Plastics Engineers reproducibility.

JOURNAL OF VINYL & ADDITIVE TECHNOLOGY—

—2010
 TABLE 1. Technical data sheet of LLDPE 0209 AA grade. TABLE 3. Technical data sheet for PEG 4000.

Test
Hydroxyl value (mg KOH/g) 25–32
Unit Typical value method Melting range
Lower value ‡588C
Physical Property
Upper value
618C
Melt flow rate g/10 min 2.2 ISO 1133 Average molecular mass (g mol21) 3500–4500
(1908C/2.16 kg) Identity (IR) Passes test
Density g/cm3 0.920 ISO 1183
Film Properties
Impact strength g 130 ASTM D1709
Tear strength TD g/25 l 300 ASTN D1922
Tear strength MD g/25 l 110 ASTN D1922
Yield Stress TD MPa 11 ISO 527
Yield Stress MD MPa 10 ISO 527
Tensile Strength TD MPa 28 ISO 527
Blowing Film
Tensile Strength MD MPa 36 ISO 527 In order to prepare a blown film sample, a film-blow-
Elongation at break TD % 800 ISO 527
Elongation at break TD % 600 ISO 527
ing extruder, Model 18016, manufactured by Collin, a
Thermal properties German company, was used. Some of its features are:
Vicat softening 8C 93 ISO 306 Screw diameter ¼ 45 mm, L/D ¼ 25, Production
temperature capacity ¼ 35 kg/h
The film thickness was 550 lm (test method, ASTM
D6988).
EXPERIMENTAL

SAMPLE PREPARATION
Materials
Poly(ethylene glycol) and the processing aid must be
In the process of polyethylene film blowing, a mixture used as a masterbatch. Therefore, 1, 3, 5, and 8% master-
of linear low-density polyethylene and low-density poly- batches were prepared from Dynamar FX5922 processing
ethylene is used. Processing of linear low-density polyeth- aid, and a 5% polyethylene glycol masterbatch was pre-
ylene would face some processing difficulties. Table 1 pared based on LLDPE grade 0209 (Amir Kabir) by
shows the technical characteristics of linear low-density means of an extruder with a ZSK25 screw operating at
polyethylene manufactured by the Amir Kabir company 600 rpm and 32% torque. Table 4 shows the temperature
(Grade 0209). Data for Dynamar FX5922 processing aid, profile.
manufactured by 3M Dyneon, and poly(ethylene glycol) By using a film blowing device, film samples of 0.06
(merck) with a molecular weight of 4000 are given in mm thickness and 250 mm width were prepared, as
Tables 2 and 3. devised by Taguchi software. Sixteen experiments (tests)
were proposed by the software based on temperature,
screw velocity, masterbatch activity, die gap diameter,
and processing aid concentration (Table 5). Table 6 shows
Extrusion the relevant temperature profiles and the temperatures that
predominated in different regions of the device, which are
Polymer processing aids must be added to the process average values.
as masterbatches because they can react with polymer
additives. A low quantity needs to be used. In order to
RESULTS AND DISCUSSION
produce masterbatch processing aids, an extruder, ZSK25,
was used having a screw with a high L/D ratio, in order In the sixteen proposed experiments, several parame-
to provide good dispersion of the processing aid in the ters such as pressure, temperature, amperage, and time to
ground matrix. The screw diameter was 25 mm, and the stabilize pressure drop were recorded. Three parameters
L/D ratio was 40. (die pressure drop, amperage, and time of changes to

TABLE 4. Temperature profile of extruder for compounding.

TABLE 2. Technical data sheet for Dynamar FX5922.
Zone Temperature (8C)
Form Powder
Color White 1 170
Activity percent 97% 2 180
Organic additives 3% 3 190
Particle size Less than 10 mesh 4 190
Density 0.7 5 190
Level of use 200–800 ppm 6 195

DOI 10.1002/vnl JOURNAL OF VINYL & ADDITIVE TECHNOLOGY—

—2010 79
 TABLE 5. Taguchi designed tests. TABLE 6. Temperature profiles (8C) of film blowing process.

Screw PPA Temperature Zone Zone Zone Zone Zone Zone Zone Zone
Die gap velocity Temperature concentration profile 1 2 3 4 5 6 7 8
(mm) (rpm) MB (%) profilea (ppm) No.
1 185 190 195 200 205 205 205 205
0.8 35 1 1 200 1 2 190 195 200 205 210 210 210 210
0.8 40 3 1 300 2 3 195 200 205 210 215 215 215 215
1.2 45 5 1 500 3 4 195 205 210 215 220 220 220 220
1.2 50 8 1 700 4
1.2 35 1 2 200 5
1.2 40 3 2 300 6
0.8 45 8 2 700 7 remain stable) were selected, and the summation results
0.8 50 5 2 500 8 were entered in the software to provide the responses
0.8 35 5 3 700 9 shown in Fig. 1. Table 7 shows the relationship between
0.8 40 8 3 500 10
1.2 45 1 3 300 11
the level of each design parameter and the corresponding
1.2 50 3 3 200 12 impact on performance variation.
1.2 35 8 4 300 13 The optimum conditions of the experiments anticipated
1.2 40 5 4 200 14 by Taguchi software are 2208C for processing tempera-
0.8 45 3 4 700 15 ture, 35 rpm for screw velocity, 5% for masterbatch activ-
0.8 50 1 4 500 16
ity, 700 ppm for processing aid concentration, and
a
See Table 6. 0.8 mm for die diameter. The algorithm was confirmed

FIG. 1. Effect of test parameters on average response. [Color figure can be viewed in the online issue,
which is available at www.interscience.wiley.com.]
 TABLE 7. Taguchi results. TABLE 9. Error in total without rpm.

Mean f SS V SS0 q
PPA Temperature 3 587.99 195.99 551.96 11.09
Temperature Screw concentration Die gap Screw velocity 0 0 0 0 0
Level (8C) velocity (rpm) MB (%) (ppm) (mm) MB 3 608.33 202.77 572.3 19.73
PPA concentration 3 2871.89 957.29 2835.86 60.10
1 29.18 33.67 24.97 23.36 46.35 Die gap 1 392.67 392.67 380.66 8.06
2 32.68 29.51 37.21 22.37 22.4
Error 5 60.04 12 0.79
3 40 39.35 41.78 37.27
Total 15 4718.04 100
4 44.66 34.98 33.56 54.54
Degree 4 5 3 1 2

In order to conduct an analysis of the relative impor-

by the results obtained when experimental melt instability tance of each factor more systematically, an analysis of
shifted toward high shear rate at high temperature. variance (ANOVA) was applied to the data. The main
Velocity gradient of the polymer had a lower slope objective of the ANOVA is to extract from the results
based on decreased molten viscosity of the polymer at how much variation each factor causes relative to the total
high temperature. Hence, the polymer processing aid variation observed in the result. The results are given in
needed a shorter time to decrease the velocity gradient. Table 8, which shows the contribution of each design
Higher shear rate resulted in melt instability. The proc- parameter to the total design-to-design variation in poly-
essing aid migrated toward the interfacial surface more mer processing aid performance. In this table, the number
rapidly and showed its effects. of degrees of freedom (DF) of a parameter is equal to the
On the basis of Fig.1, the best activity of the master- number of levels of that parameter, minus one. The sum
batch was determined to be 5%. It must be noted that the of the squares of the source, marked SS, measures the
processing aid was extruded twice, first during master- variation in terms of the sum of the squared deviations
batch production and then during film blowing. Lower-ac- from the mean. The variation of the source is marked V.
tivity masterbatches do not have a homogeneous disper- The pure variation of the source, SS’, is the variation of
sion in the first stage of masterbatch production. Higher- the parameter without the error variation and is calculated
activity masterbatches have a heterogeneous dispersion for the unpooled parameters. The percentage contribution
during film preparation because of the lower quantity of the source to the total variations is ‘‘q’’ [6]. From the
used. ANOVA results in Table 8, the polymer processing aid
An increased screw velocity leads to a lower residence concentration had the largest variance; the masterbatch
time for the polymer. This, in turn, leads to migration of percentage was second; and screw velocity had less
a processing aid to the extruder inner surface. On the involvement and could be neglected. Upon ignoring it,
other hand, fast rotation of the screw results in a higher the error became 0.79 (Table 9).
amperage, and that is worse for industrial application. As was already mentioned, several tests were run in
The results of these factors lead us to choose 35 rpm as order to find the best quantity of processing aid Dynamar
the best value. It must be noted that at high rotations, sta- FX5922. The concentrations examined (220, 250, 280,
bilization of a film bubble will lead to difficulties and will 300, and 700 ppm) are shown in Fig. 2. From the function
be associated with a high quantity of waste material.
A decreased die thickness provides an increase of shear
stress. Hence, polymer instability occurs at lower shear
rates when the die diameter is wide. Although a smaller
die gap results in a more rapid formation of a thin cover-
ing layer of processing aid, best results are found with
0.8 mm as the die gap.

TABLE 8. Error in total.

DF SS V SS0 q

Temperature 3 587.99 1958.99 497.93 10.55

Screw velocity 3 197.12 65.71 10.06 2.27
MB 3 608.33 202.77 518.27 19.58
PPA concentration 3 2781.89 957.3 2781.73 58.96
Die gap 1 392.67 392.67 362.65 7.69
Error 2 60.04 30.02 0.950
Total 15 4718.04 100
FIG. 2. Data for film blowing optimization with Dynamar FX5922.

DOI 10.1002/vnl JOURNAL OF VINYL & ADDITIVE TECHNOLOGY—

—2010 81
 TABLE 10. Pressure drop for film blowing optimization amount. TABLE 12. Impact resistance data for film blowing optimization (Test
method: ASTM D1709).
PPA concentration Pressure drop
(ppm) (bar) Sample Thickness (mm) weight (g)

220 46 LLDPE 0.06 202

250 51 PE þ 220 ppm PPA 0.065 211
280 61 PE þ 250 ppm PPA 0.061 215
300 75 PE þ 280 ppm PPA 0.0615 216
700 104 PE þ 300 ppm PPA 0.062 238
PE þ 700 ppm PPA 0.065 242

and price points of view, an acceptable quantity would be For optimizing processing conditions mixtures of
as much as 250 ppm (Table 10). Dynamar FX5922, and PEG with a molecular weight of
In industry, film thickness is used indirectly for output 4000 were used as a two-component processing aid.
determination. Film thickness is calculated by means of Results are shown in Fig. 4. The two-component process-
ASTM D6988. Then by using the mass balance rule for ing aid produced the following results:
the die exit and pulling roll DDR and BUR formulas, the
following equation is extracted [7]. 1. The PEG improved the effects of Dynamar FX5922.
2. Dynamar FX5922 needed time to show its performance,
output ¼ 23106 3 3600 3 film thickness 3 film width but PEG indicated its performance at exactly the time of
  addition.
die width
3 TOF 3 density 3 1  3. If the amount of PEG exceeded that of Dynamar FX5922,
die diameter it had less influence.
4. If the amount of Dynamar FX5922 exceeded that of PEG,
In this equation, TOF is the velocity of the drawing Dynamar FX5922 performance was enhanced.
roller (m/s), output is in kg/h, and the other parameters
are in mm. Output percentages are listed in Table 11. As mentioned above, PEG works as an external lubri-
Impact resistance tests have also been performed on cant, leads to an increased migration to the interfacial sur-
the basis of ASTM D1709 standards, and the results are face, and increases polymer bulk. It must be noted that if
listed in Table 12. The ASTM D1709 method employs a PEG content is higher than Dynamar FX5922 content, it
dart with a 38.10 6 0.13 mm [1.500 6 0.005 in.] diame- leads to an overlubrication phenomenon, the effects of
ter hemispherical head dropped from a height of 0.66 6 which are motor electric current fluctuation and pressure
0.01 m [26.0 6 0.4 in.]. This test method may be used variation. Screw and barrel slip produces these results.
for films whose impact resistances require masses of Quantitative results provided by the tests are as below:
about 50 g or less to about 2 kg in order to fracture them.
Table 12 shows that an increasing concentration of proc- 1. Adding 200 ppm of PEG to 300 ppm of Dynamar
essing aid leads to an increased thickness of film that FX5922 improves performance by about 20% as com-
improves impact resistance. Results for 300 and 700 ppm pared to that of 500 ppm of Dynamar FX5922.
are not so different. Percentage increases of resistance are 2. Adding 100 ppm or 200 ppm of PEG instead of Dynamar
15.1 and 16, respectively, and they show that 300 ppm is FX5922 weight equivalents gives performance improved
over that of 500 ppm of Dynamar FX5922 by 65.3% and
commercially preferred (Table 12).
49.4%, respectively.
Figure 3 also shows that increasing film thickness leads
to increased impact resistance resulting from ever-increas-
ing concentrations of processing aid. These results show
that not only surface smoothness but also increasing
thickness results in higher impact resistance.

TABLE 11. Output increase for film blowing optimization.

PPA concentration
(Dynamar FX5922) (ppm) Output increase (%)

220 7.69
250 3.22
280 2.43
FIG. 3. Relation between impact resistance and film thickness. [Color
300 2.12
figure can be viewed in the online issue, which is available at www.
700 1.63
interscience.wiley.com.]

82 JOURNAL OF VINYL & ADDITIVE TECHNOLOGY—

—2010 DOI 10.1002/vnl
 aid at the petrochemical research and development
company.

REFERENCES

1. S. Papp, S.E. Amos, and J. Briers, International Business Fo-

rum on Specialty Polyolefins, SPE 99, Houston (1999).
2. T.J. Blong, D.F. Klein, A.V. Pocius, and M.A. Strobel,
TAPPI Polymer, Laminations and Coatings Conference Pro-
ceedings, Atlanta (1999).
3. S.G. Hatzikiriakos and K.B. Migler, Polymer Processing Insta-
bilities: Control and Understanding, Marcel Dekker, New
York, 237 (2004).
FIG. 4. Data for PEG þ Dynamar FX5922 (PPA) mixture. [Color
figure can be viewed in the online issue, which is available at www. 4. S.G. Hatzikiriakos and K.B. Migler, Polymer Processing
interscience.wiley.com.] Instabilities: Control and Understanding, Marcel Dekker,
New York, 261 (2004).
3. As mentioned above, the best results for the mixture of
5. R. Gächter and H. Müller, Plastics Additives Handbook: Sta-
processing aids are obtained with 400 ppm of Dynamar
bilizers, Processing Aids, Plasticizers, Fillers, Reinforce-
FX5922 plus 100 ppm of PEG.
ments, Colorants for Thermoplastics, Hanser, 457 (1993).
6. R.K. Roy, Design of Experiments Using the Taguchi
ACKNOWLEDGMENTS Approach: 16 Steps to Product and Process Improvement,
Wiley-IEEE, 172 (2001).
The authors wish to express their special gratitude 7. S. Shafiee and M. Habibollahi, Film Blowing Process, Petro-
to Mr. Habibollahi and Mr. Ranji for their painstaking chemi, Tehran, 100 (2004).

DOI 10.1002/vnl JOURNAL OF VINYL & ADDITIVE TECHNOLOGY—

—2010 83