.1\

 

 

 

E4.‘-‘-' 4

gnu-u.

 

“’9 M W5 mWWW

ll , l.l’lllllll’ll/l/Ul

1 3 1293 00575 6907

 

 

 

 

 

 

I".

 

Llama?
Mkhigan State
University

 

 

 

This is to certify that the

thesis entitled
Mechanical Property Enhancement Of Recycled
High Density Polyethylene And Wood Fiber Composites
Due To The Inclusion Of Additives

presented by

Kristine . A. Nieman

has been accepted towards fulfillment
of the requirements for

Masters Science (Packaging)

degree in

 

 

J/a‘ - cg, 1/64 -

Major professor

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

     

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

{A

 

 

 

 

 

 

 

 

 

l
mm
x
n?
'MAGié 2 |

5 q
{-24,

00 .3 £739

 

 

 

 

 

 

29

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative ActloNEqual Opportunity Institution

MECHANICAL PROPERTY ENHANCEMENT OF RECYCLED
HIGH DENSITY POLYETHYLENE AND WOOD FIBER COMPOSITES
DUE TO THE INCLUSION OF ADDITIVES
BY

Kristine A. Nieman

A THESIS

Submited to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

School of Packaging

1989

ABSTRACT

MECHANICAL PROPERTY ENHANCEMENT OF RECYCLED
HIGH DENSITY POLYETHYLENE/WOOD FIBER COMPOSITES
DUE TO THE INCLUSION OF ADDITIVES

BY

Kristine Anne Nieman

Promotion of interfacial adhesion and fiber dispersion were
sought through the inclusion of additives in. high density
polyethylene (HDPE) and wood fiber composites so as to enhance
mechanical properties. Five additives were used to modify the
recycled HDPE/wood fiber composite. Specimens were tested for
tensile properties, impact strength, water sorption and creep.
Specimens were also analyzed using scanning electron microscopy
(SEM). Two of the five additives, low density polyethylene and
stearic acid, were determined ineffective for enhancing
properties. Chlorinated polyethylene had little effect, either
positive or negative, on the composite's properties. Maleic
anhydride modified polyprOpylene displayed potential for
improving adhesion between the recycled polyethylene and wood
fibers, based on improvements in tensile strength and modulus
and SEM results. Ionomer modified polyethylene also displayed

some positive results.

to my parents, for all the years of encouragement

iii

ACKNOWLEDGEMENTS

I would like to thank my major professor, Susan Selke, PhD.
(School of Packaging, Michigan State University), and my
committee :members, Jack; Giacin, PhD. (School of .Packaging,
Michigan State University), Thomas Pinnavaia, PhD. (Department
of Chemistry, Michigan State University), and Kit Yam, PhD.
(Food Science Department, Rutgers University) for their
academic support and guidence without which I could not have

completed this work.

I would also like to thank. Mike IRich from ‘the Composite
Research Center for the instruction and use of equipment, Alan
Sliker, PhD. for sharing his knowledge of wood technology, and
Varsha Kalyankar for her help with the extrusion process and
her advice. A special thanks goes to the companies that

donated materials that enabled this research to be conducted.

I would also like to express thanks to my family and friends

for their support.

iv

TABLE OF CONTENTS

page
LIST OF TABIIES ..OOOOOOOOOOOOOOOOOO ...... ..OOOOOOOOOOOOOOOVii
LIST OF FIGURES ..OOOOOOOOOOOOOOOOOOO ........ O. .......... OOViii
I. INTRODUCTION ..OOOOOOOOO... ........... O... ..... 0.0.0.1
II. EXPERIMENTAL
A. materials 000...... OOOOOOOOOOOOOOOOO 0.0.00.0000006
B. Chemical comPOSition O O O O O O O O O O O O O O O O O O O O O O O O O O .10
Co Methads OOOOOOOOOOOOOOO... ......... 0.0.00.00000013
III. LITERATURE REVIEW
A. Composite Materials .............. ......... .....19
B. Prediction of Composite Properties ........... ..20
C. Interfacial Strength ........................ ...25
D. Review of Prior Research ..... . ................. 31
IV. RESULTS AND DISCUSSION

A.

B.

C.

D.

E.

F.

Differential Scanning Calorimetry .............. 34
Tensile Properties .............. . .............. 39
Impact Strength ................................ 48
Water Sorption ............... . ............... ..51
Creep Test .................. ....... ............54
Scanning Electron Microscopy ............... ....57

VI.

VII.

Page
SUMMARY

A. summary and conCIuSions 00......0.00.00.00.0000063

B. Recommendations For Future Work ................65

APPENDICES
A. Composite Contents .............................66
Manufacturers of Materials .....................67

B. Data and Statistical Analysis ..................68

BIBLIOGRAPHY
A. List of References . ...................... . ..... 93

B. General References .............................95

vi

Table

LIST OF TABLES

Page
Additives Used ........................................7
Differential Scanning Calorimetry ....................34
Tensile Strength .....................................40
Modulus of Elasticity ........... ..... ................41
Elongation at Break ..................................42
Recycled HDPE vs Recycled HDPE + MA.PP ...............42
Izod Impact Strength ................. ............ ....48
Water Sorption ...................... ..... ............51

creepAnaIYSis .0....0.0.00.00.00.000.0.0.00000000000054

vii

LIST OF FIGURES

Figure Page

1 Differential Scanning Calorimetry Results............35

2 Tensile Strength ............ ..... ...................43
3 Modulus of Elasticity ....... ....... ................. 44
4 Elongation at Break ....... .............. .. ...... ....45
5 Izod Impact Strength ................ .......... ......49
6 Water Absorption .. ........ ......... ............... ..52
7 Creep Analysis ...................... ............. ...55

8 Scanning Electron Microscopy Results ................58

viii

I . INTRODUCTION

INTRODUCTION

The use of plastics jpackaging' is. expected to escalate as
society becomes more and more convenience and time oriented
(Melosi, 1981). Plastics are lightweight, shatterproof, and
cost effective and are rapidly replacing other packaging
materials as gains are made in plastics technology. Plastic's
expanding share of the municipal waste stream, its
nonbiodegradable characteristics and the growing shortage of
landfill space poses an eminent problem as plastic use
increases. Pdastic's share of the municipal waste is at 7.2%
and is expected to increase to 9.8% by the year 2000
(Leaversuch, 1987) . These figures are based on weight. The
magnitude of the problem becomes even more significant when
weight is converted to volume. Plastics packaging waste once
converted to volume is figured to account for 31.4% of the
materials in the municipal waste stream and is projected to be
at 37.7% by the year 2000 ("Analyst: Solid Waste Becomes
Crisis, 1988). Approximately 25% of the total packaging
market is plastic. Plastics have such desirable properties
that their use is expected to grow to 50% by the year 2000. It
is evident that the use of plastics in the packaging market has
become so prominent that banning is not plausible. Yet, in

order to continue enjoying plastic's many advantages, its

2
disadvantages must also be dealt with. In 1988, approximately
2000 bills were introduced directed at the municipal waste
problem (Serie, Mattheis, 1988). Packaging container
legislation accounted for an estimated 300 bills with
approximately 70% aimed directly at, or concerned with plastic
packaging (Serie & Mattheis,1988). Packaging legislation
includes taxes on litter stream type waste items, deposit laws,
labeling so that plastics can be easily identified and
separated to enhance recycling, regulatory review of packages
and packaging materials, and prohibitions (Wright, 1987). The
banning of plastic has been directed mainly at Polyvinyl
chloride (PVC) because of the chlorine it releases when
combusted, at plastics containing lead and cadmium because the
metals form toxic ash when combusted, and at foamed polystyrene
because it is not recyclable and emits chlorofluorocarbons

during processing (Wright, 1987).

It is felt that a combination of recycling, landfilling and
incineration would be the most effective way of dealing with
plastic waste (Schneidman, 1987). According to the
Enviromental Protection Agency, landfill numbers have dropped
50% in 1986 as compared to 1979. Present landfill space is
reaching capacity and it is thus becoming more and more
expensive. Incineration is a method for recovering energy from
solid waste. Plastics, because they are petroleum based, are a

significant contributor to the amount of recovered energy.

3
Problems have arisen concerning plastics incineration, however,
in that some individuals feel it is a dangerous pollutant
because of the toxins that various plastics emit. Thus, there
is a debate over plastics place in incineration. Methods for
incineration are also very expensive to set up and operate.
Pyrolysis, a method in which solid waste is converted into
gaseous, liquid or solid fuels by heating organic waste in an
atmosphere of low oxygen so that combustion does not occur but
chemical decomposition does, is another way to recover energy
from solid waste. This method of energy retrieval has not been
proven reliable, is far 'too expensive ‘to Ibe feasible, is
considered to be the least advanced of the energy recovery
technologies, and is therefore not considered as an effective
alternative for relieving the plastic waste problem (Melosi,
1981). Although recycling may be the one method that provides
the highest recovery value for plastic, the practice of
plastics recycling is almost nonexistent. Currently, just over
1% of all plastic packaging is recycled, according to Wayne
Pearson, executive director of the Plastics Recycling
Foundation (Schneidman, 1987). Plastic waste is a valuable
resource that has been back-shelved due to problems with
collection, identification and markets. Growing concern over
plastic consumption is likely to force the recycling dilemma on
industry and the governemt, thus it is important to study the
use of recycled plastics and their properties before social and

legal action heightens. The future of plastics packaging may

4
rest on the ability to find methods and markets for recycled

plastic.

Polyethylene terephthalate (PET) bottles are at present one of
the few polymers actively sought and recycled. About four
times as much tonnage is generated by high density polyethylene
(HDPE) , made into dairy bottles and various other containers,
than that of PET ("Milk Bottles Reembodied", 1987). HDPE use
for dairy bottles alone is equal to the entire PET bottle
market ("Milk Bottles Reembodied, 1987). HDPE is easily
identifiable and readily recyclable, due to the new washing
systems that have been developed. Markets for the recycled
HDPE are being investigated. The polymer is limited in its use
for structural applications, due to its low stiffness and high
creep properties. It is hoped that this particular drawback
can be overcome by reinforcing the polymer with a stiff and
strong filler. Reinforcing the polymer with a filler can
increase its marketability by decreasing cost, obtaining
special properties and improving load bearing capabilities.
Wood fiber has been recognized as a possible filler because of
its low cost, stiff and strong fibers, ease of processability
and its availability. Unfortunately, cellulose fibers are not
compatible with HDPE. The wood fibers are hydrophylic and
polar while the polymer is hydrophobic and nonpolar. There is
a lack of adhesion between the phases resulting in poor

mechanical properties. When two dry substances are pressed

5
together in the absence of a bond, little effort is needed to
pull them apart. Interfacial forces acting to adhere the two
phases together will increase the composite's strength. If
fibers are "wetted-out", ie, each fiber is totally enclosed by
the matrix, and better dispersion of the fibers is achieved,
improved mechanical properties will result. Prior work done in
the area of short fiber reinforced thermoplastics has shown
that cellulose fibers have not resulted in any significant
degree of reinforcement, despite their stiffness and strength
properties (Klason et al., 1984). The reason for this is
thought to be the result of fiber damage occurring during
compounding and processing and a lack of adhesion between the
phases (Klason et al., 1984). The fiber stiffness and strength
can be taken advantage of, if adhesion between the phases can

offset some of the strength lost due to fiber damage.

The primary objectives of this investigation were to: (i)study
the fiber-matrix interface of a recycled HDPE and wood fiber
composite ; and (ii) develop a method to achieve good fiber
dispersion and adhesion between the phases so as to obtain a

strong composite, and a viable recycled material.

II . EXPERIMENTAL

MATERIALS

The materials used to make the composites for this study

consist of the following.

\A). High Density Polyethylene dairy bottles were collected,
cleaned using water, and the labels and caps removed before
granulating into resin using a Lowline Granulator Model 68-913
from Polymer Machinery Corp. The resin supply used for this
study was approximately 20% recycled unused HDPE and 80%
recycled used HDPE. The dairy bottles were collected from
several different dairies. Virgin and recycled HDPE resin
samples were characterized using differential scanning
calorimetry to determine changes in crystallinity and melt
temperature. Virgin HDPE "Fortiflex A60-70-119" from Soltex

Polymer Co. was used for purging during the extrusion process

and the DSC test.

{I B. Aspen hardwood fibers obtained from Canfor Canadian Forest
‘Products, were used as a reinforcing filler. Fiber bundles
are formed by mixing wood chips and shavings with steam under
pressure and refined using electric motors and refiner plates
(rotating disks). A high yeild is acheived with very little
damage occurring to change the lignin or hemicellulose. Aspen
wood fiber cost is approximately $0.10/lb including freight

cost. Cellulose is a hydrophilic glucan polymer. Most

7
hardwood species contain four types of cells; vessel segments,
fibers, and transverse and axial parenchyma. The fibers
perform the support role. Fibers are thick-walled, elongated
cells with closed pointed ends. Fibers range in length from
.7m to 3mm.(Goldstein, 1977). The large amount of hydroxyl
groups that occur throughout the structure can attract and hold
water molecules by hydrogen bonding. Before the fibers are
incorporated into the composite they are removed from their

container and allowed to air dry for two to three days.

C. Five additives were studied for’ their effect, on the
recycled HDPE/wood fiber composite. They are listed in Table 1

followed by a description of each.

 

Table 1
Additize§_used

1. Chlorinated Polyethylene (CPE 4213, 40% chlorine, DOW)
Cost = $0.89/1b/truckload.

2. Ionomer Modified Polyethylene (Surlyn 1605, Du Pont)
Cost = $1.27/lb/truckload.

3. Low Density Polyethylene (LDPE, DOW)
Cost = $0.58-$0.64/lb/truckload or $0.53-$0.58/lb/railcar.

4. Maleic Anhydride Modified Polypropylene (Hercoprime,
Himont)
Cost = $12.00/lb.

5. Stearic Acid (Sigma)
Cost = $1.12 - $1.84/gram (price varies with quantity
purchased).

 

8
1. Chlorinated polyethylene (CPE) has excellent elongation and
impact resistance, but poor creep resistance. CPE was selected
for its polar nature which may aid in interfacial bonding when
added to the composite. Maximum softness for CPE is obtained
with 35-40% chlorine (Herman et al., 1981). The CPE used for
testing ‘with. the recycled, HDPE/wood fiber composite ‘has a
chlorine content of 40% by weight. CPE can cause minor eye and
skin irritation due to the evolution of HCL at high
temperatures. It can not be used to package fatty or oily

foods.

2. Ionomer modified polyethylene is a thermoplastic material
was also selected because of its polar nature. Ionomer
modified polyethylene has ionic and polar bonds which may aid
in interfacial bonding. Ionomers are transparent, tough,
flexible, and have good abrasion resistance and excellent
filler acceptance. It adheres to metals, nylon, other

polyolefins and urethane finishes.

3. Low density polyethylene (LDPE) due to its many side
branches has low crystallinity, is flexible and translucent,
and has excellent impact resistance. Addition of LDPE to HDPE
will result in decreasing percent crystallinity and viscosity
during processing, thus decreasing brittleness of the final

product.

9
4. Maleic anhydride modified polypropylene (MA.PP), a coupling
agent, is the fourth additve listed. A coupling agent is one
way to improve adhesion between the two phases. It acts as a
link between the fiber and the matrix, thus the composite's
strength will improve with bonding of the fiber-matrix.
Covalently bonded materials form a structure that acts as one
unit. Without the bond the two phases are only blended

together and can be easily pulled apart.

5. Stearic acid is to be tested as a dispersant. It was
selected for study based on the potential for preventing
agglomeration of the filler particles. Evenly dispersed filler
throughout the matrix will increase the mechanical properties
of the composite. Agglomeration of the filler weakens the
structure by causing points that will fail under stress. To
maximize strength each fiber should be completely enclosed by
the matrix (Folkes, 1982). Stearic acid is hoped to enhance

the composite's morphology and strength.

CHEMICAL COMPOSITION
High Density Polyethylene (HDPE)

{C82 — CH2 - CH2 - CH2 - CH2 - CH2 - CH2 }

 

 

 

 

n

0 65-90% crystalline

o 130-1350c Tm

o .94-.9659/cc density

0 ~120°c T9

0 hydrophobic, nonpolar

Aspen Hardwood Fibers
cnzou o H OH

O H H OH H O

O

H OH H H H H

H on CHZOH o

o hydrophylic, polar
0 very crystalline
0 cell wall: cellulose (40-60%)

lignin (20-30%)

10

11
C. Additives

1. Chlorinated Polyethylene (CPE)

" CH2CH2CH2 " + C12 —'—9 [ CflszCHz} 4" BC].

c1 n

2. Low Density Polyethylene (LDPE)

- CH2 - CH2 - CH2
- CH2 - éH - CH2 - CH2 - CH - CH2
4...
AH - CH2 - CH2 - CH2
- A...

0 40-60% crystallinity
o 105-1150c Tm

o .916-.932g/cc density

3. Ionomer Modified Polyethylene

CH3 CH3
l I
CH2 = c + CH2 = CH2 ———) - CH2 - CH2 - CH2 - c - CH2
I l
COZH CO2H

12

4. Stearic Acid

0
I

CH3 - (CH2)16 - C - OH

0 70°C Tm

 

 

5. Maleic Anhydride Modified Polypropylene (MA.PP)
O O O O O O
\/\// \\/\/
T ‘f EH3 .6 i *r
{ CH2 - f f ‘ CH2 ‘ CH2 ‘ f ’ CH2 ‘ C C ‘ CH2 - CE} n
l I
H H

H H H CH3

METHODS

W

The recycled HDPE is prepared as stated in the HDPE materials
section. Each composite, with a few exceptions, is comprised
of 30% wood fibers, 5% additive and 65% recycled HDPE by
weight. (See appendix A for a detailed breakdown of the
composite contents.) The recycled HDPE/wood fiber composites,
with various additives, are produced utilizing a Baker Perkin,
Model MPC/V-30 DE, 38 mm, 13:1 co-rotating twin screw extruder.
The additives were mixed with the polymer prior to being added
to the extruder's hopper. The mixing was done by thoroughly
shaking the two in an enclosed container. The polymer was
premelted in zone one and the wood fibers were hand fed into
the extruder at zone 2. Adding fibers to pre-melted polymer is
thought to be advantageous in reducing fiber damage and in
gaining better dispersion. All three extruder zones and the
die were preheated and maintained at a temperature of 150°C.
The compounder speed was set at 150 rpm's. Feed rate of the
polymer can be varied with the desired percent wood fiber. For
30% wood fiber the rate was 4 (Feed rate setting is based on %

of compounder speed rate).

The extruded composite rods are then converted into sheets
using the Carver laboratory press compression molding machine,

model M25 ton. The upper and lower platens are maintained at

13

l4
150°C. An initial ten minute warm up period is followed by ten
minutes with the pressure held at 30,000 psi or more. The
platens are then. water’ cooled. to room. temperature for 15
minutes before the sheet is removed. .Approximately' three

sheets can be made from 350 grams of material.

In order to evaluate the properties of the composites, the
following ASTM standards were employed. The composites were
initially screened to determine the effectiveness of the
additive, using tensile and impact testing. If results were
positive, further testing was conducted in the areas of water

absorption, creep, and Scanning Electron Microscopy.

A. Ignailg Ezgngzfix Qgtgzminggign
o ASTM standard D638-77a Tensile Properties of Plastics.

0 Equipment: Instron, model 1114

Tensilkut cutting machine

Dumbbell-shaped Type I specimens are cut from sheets such that
they were .5 inches at the narrow section, using the Tensilkut
cutting machine according to the ASTM standard. Specimens were
tested on the Instron with a full scale load of 500 lbs, chart
speed at 10 in/min., and crosshead speed equal to .5 in/min.
The specimens were conditioned prior to testing at 23 +/- 2°C

and 50 +/- 5% RH for not less than 40 hours. Abrasive paper

15
is used to keep the specimen from slipping in the grips.
Specimens that did not break in the narrow section were
disregarded. Tensile strength, elongation at break and modulus
of elasticity are calculated from the chart recorder results,

using the following formulas.

 

(1)
AL - Ax sshea eed n
chart speed (in/min)
where: 01- = change in gage length
AX a distance traveled on the chart
Gage length = length between grips
For Modulus: Stress = uforce (2)
original minimum cross-sectional area
Strain = cha e ' n st e ch , (3)

original gage length

Tensile Strength = highest stress a material can carry

% Elongation at Break = strain at break x 100

B. Izod Impact Strength Determination
o ASTM standard 0256-81 Impact Resistance of Plastic and
Electrical Insulating Materials.

0 Equipment: TMI 43-1 Izod Impact Tester

16

Specimen notcher

Test specimen are cut to the standard width of 0.5 inches (1.27
cm) and are 2 inches (5.08 cm) in length from compression
molded plates. The samples are conditioned prior to testing
for not less than 40 hours at 23 +/- 2°C and 50 +/- 5%RH. A
five pound pendulum is used for the test. According to the
Izod impact test requirement, the specimen is notched. The
notch allows for a brittle rather than a ductile fracture.

Values are stated in inch-pound units.

Win.
0 ASTM standard D 570 Water Absorption of Plastics.

0 Equipment: Instron, model 1114

Samples are tested for dimensional stability using the long

term immersion method. Moisture gain over time is measured.

WW
0 ASTM standard D 2990-77 Tensile, Compressive, and
Flexural Creep and Creep-Rupture of plastics.

0 Equipment: clamps, fifty pound weights

Creep test provides information that will aid in predicting the

17
strength of a material subjected to long term loads. It also
shows dimensional changes that are a result of a long term

loads.

WW

0 SEM manual

0 Equipment: SEM, model JEOL T-330

SEM aids in determining the presence of a bond and its
effectiveness. The fracture surface of a specimen is studied.
Alan Sliker, PhD. , wood scientist, Michigan State Forestry

Department, was consulted to evaluate SEM resutls.

stm
o Du Pont DSC manual

0 Equipment: Du Pont 9900 DSC

The test is utilized to determine the melting temperature and
degree of crystallinity of HDPE: virgin, used recycled, and
unused recycled. Polymer crystallinity is an indication of
strength. The more dense the polymer structure, the better the
mechanical properties of the polymer. A sample size of 10mg is
used for each test. The samples are ramped at SOC/min and a
sweep of 120°C (30°C to 150°C) is made, with each test taking a

total of 24 minutes.

18

% Crystallinity = -Afl£——- x 130 (4)
AH*f
where: ME a heat of fusion of test sample
Aflatf = a known heat of fusion of a hypothetical
100% crystalline sample. (for PE Afl*f a 68.4

calories/gram)(Brennan, 1978).

III . LITERATURE REVIEW

COMPOSITE MATERIALS

Composites consist of one or more discontinuous phases enclosed
in a continuous phase. The discontinuous phase is that which
is harder and stronger, and thus provides the reinforcement.
The continuous phase is called the matrix. The matrix keeps
the fibers separated from one another, yet holds the fibers
together while maintaining fiber orientation. It also protects
the filler from the harmful effects of the environment and from
abrasion. The matrix is a minor strengthening factor. The
main purpose of the matrix is to transmit load to the fibers,
which contribute the greater pertion of the composite's
strength. Therefore, crucial to the composite's ultimate
properties is the fiber-matrix interface. Good adhesion
between the phases is necessary for stress transfer to occur
from the matrix to the stronger discontinuous phase. Poor
adhesion will likely result in the interface being the point of
failure. Composite materials are classified as particulate or
fibrous, based on the discontinuous phase particle size and
shape. The length (2) and diameter (d ) , or the fiber-aspect
ratio ( 1/d ) of a filler particle greatly influences the
composite properties. Fibrous reinforcments have one long
dimension, whereas particulate reinforcing fillers do not. The
particulate reinforced composite will gain in stiffness but not
in strength. Fibrous reinforcements improve stiffness,

strength, and creep, all three of which are thermoplastic

19

20
physical properties in need of improvment for use in structural
applications. There are two types of fibrous reinforced
composite materials; continuous which have long fibers, and
discontinuous which have short fibers. Interfacial adhesion
between the phases is especially important for discontinuous
fiber composites. Studies have shown that the presence of
fiber ends within the body of the composite can cause crack
initiations and thus lead to potential composite failure
(Folkes, 1982). Interfacial strength will affect the
generation of microcracks at the fiber ends caused when stress
is applied. When a strong bond is present between the phases,

the cracks will not be produced along the length of the fibers.

Theories have been developed for the prediction of tensile
strength, tensile modulus and impact strength for fiber
reinforced thermoplastics. Much work has been devoted to the
prediction of tensile “prépgrties for composite materials.
Strength and toughness are more difficult to predict. The rule
of mixtures can be used. to (predict. a composite's tensile
modulus and tensile strength. For long fiber reinforced

thermoplastics, the calculation is much simpler. It is assumed

that all fibers are working at maximum efficiency and the

21
tensile force acting on the continuous reinforcement is shared
between the matrix and. all the fibers, with. the ultimate
tensile strain being reached in the fiber. It is also assumed

that the bond between the fiber and the matrix is very good.

E - E
c f”: + Eng... (5)
ac ' ”far + ”QO (6)
Where: 1; = tensile modulus

a = tensile strength

a = volume fraction
subscripts c = composite
f = fiber
m = matrix (Clegg & Collyer, 1986)

The predicted values given by equations (5) and (6), tend to be
higher than actual values. The equations are not totally
valid, especially equation (6), since additional stresses are
present, which are not considered in the rule of mixtures
equations (Clegg & Collyer, 1986).

M”
It is difficult to predict the. properties of short fiber

reinforced thermoplastics, as compared to long fiber reinforced

22
thermoplastics due to the fact that short fiber reinforcment
generally has a three dimensional distribution of fiber
orientations and a variety of fiberfilengths that result from
processing. The influence of fiber ends is to lower the
elastic modulus and strength of short fiber reinforced
composites (Agarwal & Broutman, 1980) . When predicting the
tensile modulus for short fiber reinforced composites,
additional factors must be considered. Stiffness of short fiber
reinforced thermoplastics depends on fiber .length (and/or
dispersion), volume fraction of (fibers, the stress transfer
efficiency' of the interface and fiber’ orientation (Folkes,
1982). During processing, fiber damage may occur which often

results in lower fiber aspect ratios. The term n1 can be added

as the length correction factor such that equation (5) becomes,
Ec - nIEfof + Emom (7)

The theory utilizing the length correction factor was developed

by Cox (1952) where:

.. - [1 Ml
/ (8)

where: 1 = fiber length

5- [ 26m ]1/2
EfAflnm/r) (9)

where: <3 = shear modulus of the matrix

 

radius of the fiber

r

23 .

R = mean separation of the fibers
normal to their length.

“T a the cross-sectional area of all

the fibers in the composite.

Equation (8) accounts for length variation. However, tensile
modulus depends on the fiber aspect ratio (g/d) and not just on
fiber length. A number-average fiber length must be obtained

to account for the ifii variation.

Tensile strength is dependent on fiber length, volume fraction
of fibers, the interfacial shear strength and fiber
orientation (Clegg & Collyer, 1986) . For short fibers, the
average tensile stress on the composite will be given by:
(10)
ac'-OEQn+aE%E

2
where: 3f = average fiber stress = €25] ”food"
0

If tensile stress builds up from the fiber ends in a non-linear

way then

1

of - of” 1-(1-5):c for z >2c
(11)

where:¢na= tensile stress in a continuous fiber in same

matrix under the same loading conditions.

24
of, = average stress in the discontinuous fiber

within a distance 212 of either end.

a.
ll

critical fiber length.

The fibers can be stressed to their tensile strengths when the
fiber length is greater than the critical fiber length. If it
is assumed that the fiber failure occurs when 0f - of” , then

substituting in equation (10) gives

_ , (12)
ac off-(149) 2c] +am gm

2

Comparison of equation (5) to equation (12) shows that
discontinuous fibers provide less strength than continuous
ones. If fibers are present in the matrix with lengths shorter
than the critical fiber length, they will not be capable of

supporting the load and failure will occur at the interface.

It is very difficult to predict the impact resistance of short
fiber reinforced composites. Presently there are no models
available on which to base predictions. If brittle fibers are
added to a ductile matrix, the impact strength of the composite
decreases rapidly as the fiber concentration increases (Clegg &
Collyer, 1986). This is because the matrix is confined by the
fibers and cannot deform to absorb the energy of impact. The

work of fracture depends on the ability of a material to

25
transfer stress throughout its structure. Theories conflict as
to whether adhesion has a positive effect on the impact
strength of composite materials. One theory determined that
impact strength cannot be used as an indication of adhesion
between the phases because of other factors that effect impact
resistance and speculated that a weak interface would be
essential for absorbing the energy of impact (Clegg & Collyer,
1986) . Another theory states that adhesion enhances impact
strength by allowing stress to be transferred to the fibers so
that the impact is spread over a larger area (Katz & Milewski,

1987).

Interfacial_§trength

Each fiber-matrix system has an interface unique to it. The
interface is dependent on the fiber's atomic arrangement and
chemical properties and on the matrix's molecular makeup and
chemical constitution. There are five main mechanisms that are
often used to produce a bond between two substances (Hull,
1981). Adhesion can occur at the interface with the aid of one
or more of the following mechanisms. The first is
interdiffusion, in which a bond is formed by molecular
entanglement. The strength of the bond is dependent on the
degree of entanglement and the number of molecules involved.

Electrostatic attraction between two surfaces can be utilized

26
to form a bond. The strength of the interface will depend on
the charge density. A third method used to induce bonding is
adsorption and wetting. Strong adhesion occurs only if the
entire surface of the filler is completely wetted out. Another
method, chemical bonding is done by forming a covalent bond
between compatible chemical groups on the fiber surface and in
the matrix. Interfacial strength will be dependent on the
number and type of bonds formed. Failure at the interface will
involve the breaking of bonds. The fifth method for bond
forming is mechanical adhesion in which some bonding may occur
purely by the mechanical interlocking of two surfaces.
3

There are three possible modes of composite failure. It is
often difficult to determine where the failure has occurred.
One failure type occurs at the interface with the separation of
the two phases, which would be an adhesive failure. Separation
of the fibers from the matrix is referred to as debonding.
Failure can also be cohesive in which case either the fiber
fractures or the matrix does. The type of failure is directly
related to the bond strength. It is important to be able to
measure the bond strength between the fibers and the matrix for
evaluation of the composite for end usage. Unfortunately,
satisfactory methods for measuring bond strength are not
available due to the high degree of precision required for
testing, and. because of inherent. problems with. wood fiber

specimen preparation. One of the better established tests is

27

discussed below.

Bond strength can be determined by performing tests with single
fibers. The single fiber test can give data on shear strength
of the interface bond. It has been determined that the
relationship between compressive stress are and shear stress rs

is given by: (Bull, 1981)

(13)

’3 ’3 2.50

A value for the applied compressive stress, “a , at which
debonding is first detected at the fiber ends can be obtained
experimentally in order to determine the shear strength of the

interface.

The tensile strength of the interface 'can also be determined

utilizing the single fiber test. The following formula is

 

used:
01 - ac(um - Vf)Ef
(14)
- 2
(l-i-uf Zuf )Em
where:¢q_ = stress perpendicular to the fibers

ct = net section compressive stress (load divided by

minimum area)

t
u

Poisson's ratio of the matrix

Poisson's ratio of the fiber

K
H.
u

E = Young's modulus (Hull, 1981)

28

The tensile strength of the interface is obtained fromac at

which debonding occurs.

The appearance of the fracture surface can sometimes be
utilized as an indirect measure of the strength of the
interface bond. There are generally changes in the appearance
of the matrix fracture surface that correspond to the degree of

adhesion (Hull, 1981).

The structure and properties of the fiber-matrix interface are
a major factor in the mechanical and physical properties of
composite materials. A composite with a weak interface will
have a relatively low strength and stiffness but high impact
strength, whereas a composite having a strong interface will
have strength and stiffness but is 'very’ brittle (Clegg &
Collyer, 1986). As stated earlier, a strong interface is
crucial for the occurance of stress transfer from the matrix to
the fibers. Load is transferred from the matrix to the fibers
through the fiber ends and through the cylindrical surface of

?
the fiber near the ends. '13ng continuous fibers the fiber

 

Ength-i§.-.__9?9§§3€ than the length over whichwtflhe transfer of

_-—“-.—~—- ”—9—-

 

 

stressw occurs and the effect of the fiber ends can be

 

dismissed. This cannot be done for short fiber reinforced

 

._-.._v i.

 

composites. The composite properties are directly related to

fiber length. Stress transfer for discontinuous short fibers

m-.4‘-, ..- -

29
is analyzed by considering the equilibrium of a small element

of fiber such that

(«1:90f + (2x:dz)r - («r2)(af + dot.)
which equals:

dof Zr (15)

 

dz (Agarwal & Broutman, 1980)

where: :r = fiber radius
7 = shear stress on the cylindrical
fiber-matrix interface.

dz = infintesimal fiber length.

Equation (15) indicates that for a fiber of uniform radius, the
fiber stress increase rate is proportional to the shear stress
at the interface. Thus fiber stress at cross-sectional
distance 2 from the fiber end can be determined by integrating
equation (15). (Agarwal & Broutman, 1980).

(16)

I
_ rdz
of afo‘i'l“.

r o

where: af‘> = stress on fiber ends.

Maximum fiber stress occurs at the midfibermlength for short
-Mwlwwwwwl.r..r -Mlmflmm “all W_ -l*
fiber (Agarwal & Broutman, 1980). \\\slmaf ~““

 

The minimum fiber length in which the maximum fiber stress can

be achieved is defined as a load transfer or the critical fiber

30
length. It is over this fiber length that the load is

transferred from matrix to fiber.

2c "in
:1— - Zr
Y (17)

where: 1c = critical fiber length
d = fiber diameter
9&1: maximum allowable fiber stress (or the fiber
ultimate strength)

Y'= matrix yield stress in shear

The critical length is sometimes referred to as the
'ineffective length' because over this length the fiber
supports a stress less than the maximum fiber stress. Shear
stress ( r ) depends on processing conditions and interfacial
adhesion. If adhesion between the phases is strong, then
shorter fibers can be used to effectively reinforce the matrix

(Katz & Milewski, 1987).

31
WW

Polymers reinforced with cellulose fibers have been researched
by many to determine the effect of cellulose as a
reinforcement . Following is a summary of some of the work

completed in this area.

Sanschagrin, Sean and Kokta (1988) studied the encapsulation of
cellulose fibers mixed with polystyrene at various
concentrations. They compared the mechanical properties
determined experimentally with theoretical predictions. It was
concluded that the large differences occurring between
experimental and calculated values are due to factors such as
fiber orientation and fiber aspect ratio, which are not
accounted. for in the theoretical predictions. It was
determined that mechanical properties improved with the
reinforcement for oriented composites but a coupling agent was
needed for property enhancement with an unoriented composite

(Sanachagrin et al., 1988).

Mitchell, Vaughan and Willis (1976) studied laminates of paper
and high density polyethylene versus glass-filled high density
polyethylene for mechanical properties. They concluded that
the cellulose filled laminate compared well with the glass
filled laminate for mechanical properties yet, the full

potential of a cellulose reinforced laminate would be realized

32
only if the cellulose fibers were distributed uniformly in the
polyethylene, and if bonding were enhanced (Michell et al. ,
1976). They also concluded that water or humidity resistance
of cellulose reinforced polyethylene laminates could be
increased by acetylation or crosslinking with formaldehyde of
the fibers, although some property loss still occurred (Michell

et al., 1978).

Aspen wood fibers in the form of chemithermomechanical
pulp(CTMP) utilized as a reinformcement in polyethylene was
studied by Beshay, Kokta and Deneault (1986) to determine the
effect on. mechanical properties“ The aspen fibers showed
better mechanical properties than either mica or glass
reinforced polyethylene, and the aspen fibers improved
polyethylene's overall properties. Beshay (1986) also studied
the effect of immersion in boiling water on the mechanical
properties of a composite of linear low density polyethylene
reinforced with CTMP. The mechanical properties did not change
significantly but the fibers did improve polyethylene's
properties. The CTMP filled composite displayed better
properties than glass fiber or mica filled composites (Kokta et

al., 1986).

Zadorecki and Flodin (1986) studied unsaturated polyesters
reinforced with cellulose fibers. The cellulose fibers

increased the tensile strength and modulus of the polyester.

33
When exposed to water however, properties were lowered due to
the high amount of water uptake (Zadorecki & Flodin, 1986). It
was determined that the adhesion between the phases was not
strong during wet conditions. Formaldehyde and di-
methylolmelamine were studied for their effect on the
composite's properties when exposed to wet conditions. Water
uptake was reduced and properties improved (Hua et al., 1987).
Hua, Flodin and Ronnhult (1987) also studied mono- or di-
methylolmelamine(DMM) resin treated cellulose for their effect
on reducing water absorption. wet strength of the cellulose-

polyester composite improved considerably (Hua et al., 1987).

High density polyethylene filled with cellulose-based
reinforcements was studied by Klason, Kubat and Stromvall
(1984). These authors determined that the cellulose fibers did
not produce any significant degree of reinforcement for the
composite. Fiber damage occurring during compounding, poor
fiber dispersion and poor adhesion between the phases were
determined to be the reasons for the lack of property
enhancement (Klason et al., 1984). A second study was
conducted to determine if the above mentioned problems could be
overcome with the inclusion of additives in the composite.
Some of the additives that were chosen as dispersion aids did
help in promoting better dispersion of the fibers. However,
only one additive was found to induce adhesion namely, maleic

anhydride modified polypropylene (Dalvag et al., 1985).

IV. RESULTS AND DISCUSSION

34
A. Diffsrential_Scannins.§algrimetrx_12§§l

1. Results:

The average percent crystallinity and melt temperature (Tm) of
virgin high density polyethylene, as compared to recycled used
and unused high density polyethylene, were determined not to be
significantly different. (see appendix B for t-test results).
Table 2 presents the data obtained from tests utilizing
Differential Scanning Calorimetry. Averages were determined
from two replications of each sample. As shown, there is only
slight variation in the results obtained. The thermograms and
accompaning DSC data of the representative samples are

presented in Figure 1, A-F.

 

Table 2

Differential Scanning Calorimetry

 

__Matgriall % Crystallinity Tm(°CL
HDPE--Virgin
Run 1 63.60 132.26
Run 2 61.27 3 .98
Average 62.80 132.12
HDPE-~Recycled, used
Run 1 62.30 132.26
Run 2 sale; Illegé
Average 62.40 132.16
HDPE--Recycled, unused
Run 1 63.79 132.18
Run 2 fillié lillfig

Average 63.70 131.99

 

Heat Flow (ll/ii

(ii/0)

Heat Flow

DIFFERENTIAL SCANNING CALORIMETRY RESULTS
Figure 1

Slllpll: HDPE VIRGIN D S C File: A: 00501.05

Size: 10.1000 e9 Operetor: K. NIEMAN
Method: eo'c T0 are RAMP a'c/m nun Dete: 04/04/08 :2: 20
Coeegnt: so cc/etn Na.

 

 

 

 
 
     

 

 

 

 

 

 

 

 

o."
+ I l L
125.74%
-0.5- 182.1.1/9
1
-1.0-‘
d
«.54
-2.0~
J
132.26’0
i
-205 F V ' l V I f v v I
90 100 110 120 130 140 150
TIlDer-eture (’Cl Senerel V2.2A DuPont 9900
Senate: HOPE VIRGIN D S C File: A: 000301.08
3120: 11.3000 e9 Operetor': K. NIEMAN
Method: SO’C T0 150.0 M B‘C/MN Rm Date: 04/04/88 12: 47
Coeeent: 50 cc/etn N2.
4; I :
125.62‘6
«0.5-1 177.401/9
J
-1.¢I-J
4
-L54
4.0-1
4 131.98‘C
-a.s . , - , - , , r .1, .
90 100 110 120 130 140 150
Temperature (’C! General v2.2a DuPont 9900

35

 

 

 

 

 

 

 

 

 

 

 

 
   

 

 

Figure 1 (cont.) 36
300010: HDPE Rec. 0350 D S C F110: 4:KNosc1.o1
C . 5120: 10.0000 e9 Operator: K. meme
Method: eo'c to 150'c RAMP S'C/KN Run Date: 04/04/00 10:57
Comment: 50 cc/ein N2.
1 ' 125.29'c ‘
70.3J/0
-0.0-
E -1.0-
El
8
z i
0.
‘3 1 al
a O
.l
-2.04
4 r
'2.5 V I f r v r I
so 100 110 120 130 140 150
Tenn-reture (’Ci Generel v2.24 DuPont 9900
Seeole: HOPE sec. 0500 D S C File: 4:KN05C1.02
D. 5120: 10.4000 no Operator: K. NIEHAN
Method: so'c TO 150°C RAMP S’C/MN nun Date: 04/04/00 11:10
Comment: 50 cc/etn N2.
V ' 125.43'0 '
79.0J/0
-0.5J
l
3? -1.0-
5i
3 1
F0
u.
t:
. «-1.5-
I
i
-2.0d
4..
132.00‘c
-205 v r v I I Y
90 100 110 120 130 140 150

Temperature (’6) Generel v2.24 DuPont 9900

Figure 1 (cont.)

E.

Heat F100 (ii/0)

H00: F100 lN/g)

Sample: HOPE REC. UNUSED

5120: 10.3000 e9

Method: 90‘0 r0 150’0 RAMP S‘CINN
Coeeent: 50 cc/etn N2.

37

DSC

F110: A:KN0501.03
Operator: K. NIENAN
Run Date: 04/04/00 11:41

 

 

 

 

 

 

 

1 1 *j__ll i
F 125.30%
-0.s-l 102.0.1/0
J
-‘e°-J
J
-1.aJ
J
-a.od
1
132.13'0
.205 ' V V I T 1 V V
90 . 100 110 120 130 140 150

50.010: HOPE REC. UNUSED

5120: 10.5000 .9

Method: 90'0 T0 150’0 RAMP S'C/NN
Comment: 50 cc/mtn N2.

Temperature ('0)

DSC

General V2.2A DuPont 9900

F110: A:KNOSC1.04
Operator: K. NIENAN
Run Date: 04/04/00 12:00

 

 

 

 

 

0.0
. ; 3
///7f 125.13'0
-0.54 102.04/0
d
'1.04
4
--1.5--1
-2.0-J
131.80'0
“2'5 f 1 r 1 r 1 v I '
90 100 110 120 130 140 150

Teepereture ('C)

General v2.24 DuPont

 

9900

38

2. Discussion:

Polymer crystallinity is one indication of the polymer's
strength. Comparison of the recycled resin with virgin resin
can give an indication of changes that may have occurred as a
result of recycling. A change in the melt temperature is an
indication that changes in crystallinity and/or molecular
weight distribution have occurred. Melt temperature is related
to processibility and flow characteristics of resin. The Tm of
the recycled resins are essentially the same as that of virgin
HDPE. A breakdown in the polymer as a result of the recycling
process would show itself in the polymer's structural
regularity (Pattanakul, 1987) . Polymer degradation, due to
initial processing and forming, consumer use, exposure and
reprocessing would be evident in a change in the resin's
properties. Prior work done by Pattanakul (1987), also
determined that there is little difference in melt flow index,
tensile strength, elongation at yield and modulus of elasticity

for recycled HDPE from milk bottles as compared to virgin HDPE.

39
B. W

1. Results:

Tensile strength results are tabulated in Table 3 and presented
graphically in Figure 2. As shown, the addition of 30% wood
fibers to recycled HDPE decreased tensile strength by
approximately 20%. Significant differences were found for
three of the seven specimens tested, namely: 5% maleic
anhydride modified polypropylene, stearic acid, and 100%
recycled HDPE when compared to the composite with no additive.
(see t-test results in appendix B). As can be seen from Table
3 and. Figure 2, incorporation. of’ stearic .acid. resulted. in
significant lowering of the composite's tensile strength, thus
having a negative effect on tensile strength. The composite
containing 5% MA.PP -had an average tensile strength almost
equal to that of the recycled resin and surpassed that of the
composite with no additive. The effect of MA.PP (5%) on
tensile strength was determined to be significantly different
than the composite with no additive, at an alpha level of .05.
Composites with 2% MA.PP, chlorinated polyethylene and ionomer
modified polyethylene also performed on the average better than
the composite with no additive, but not at a statistically

significant level.

40

 

 

Table 3
Tensile Strength
(psi)
410mm Mean so
No Additive 3914.48 378.27
Rec. HDPE (100%) 4977.62 187.75
CPE 4105.83 718.11
Ionomer 4121.72 445.95
LDPE 3555.94 690.57
MA.PP (2%) 4532.80 1003.91
MA.PP (5%) 4839.80 571.40
Stearic Acid 3134.88 555.61

 

Tensile modulus results are summarized in Table 4 and presented
graphically in Figure 3. As shown, the inclusion of 30% wood
fibers to the recycled HDPE resulted in an increase in the
modulus of approximately 65%, as compared to the resin alone.
Stearic acid and low density polyethylene's inclusion in the
composite resulted in a decrease of modulus. The composite
containing chlorinated polyethylene resulted in a modulus that
was essentially equal to that of the composite with no
additive. An increase in tensile modulus was also observed for
composites containing MA.PP and ionomer modified polyethylene,
as compared to the composite with no additive. Statistical
analysis of the data indicated that the composite with no
additive included, when compared to those with additives, is

not significantly different at a .05 alpha level.

41

 

 

 

Table 4
Modulus of Elasticity
(psi)
Material Mean. so
NO Additive 176509.30 45427.80
Rec. HDPE (100%) 111723.00 8669.82
CPE 178933.70 17348.00
Ionomer 212579.20 29186.20
LDPE 145606.90 42474.00
MA.PP (2%) 205640.50 29156.10
MA.PP (5%) 166571.40 32211.60
Stearic Acid 146164.50 26346.60

 

Elongation. at break data is presented in. Table 5 and is
illustrated in Figure 4. As shown, the inclusion of 30% wood
fiber in the recycled HDPE resulted in a substantial decrease
in elongation of the composite, as compared to recycled HDPE
resin with no fibers or additives. All composites containing
additives exhibited higher percent elongation than the
composite with no additive present. Statistical analysis
indicates that the increase in elongation is significant for
all composites with additives. (see appendix B for t-test

results).

42

 

Table 5

Elongation at Break

(%)

 

 

Material Mean SD
No Additive 1.40 .37
Rec. HDPE(100%) 240.31 70.32
CPE 3.05 .66
Ionomer 3.10 .69
LDPE 3.75 .68
MA.PP (2%) 2.72 .93
MA.PP (5%) 3.75 .63
Stearic Acid 4.08 .71

 

Specimens containing 5% MA.PP and

95% recycled HDPE were

compared with specimens consisting of 100% recycled resin. No

significant difference was found between the two for tensile

strength or modulus.

(See Table 6 and Appendix B).

 

Table 6

Recycled HDPE vs Recycled HDPE + MA.PP

Material

Tensile Strength

Tensile Modulus

 

(psi) (psi)
.Mean SD Mean SD
Rec. HDPE (100%) 4978.90 187.25 111723.00 8669.82
4730.44 286.53 129720.40 26102.30

Rec. HDPE (95%)
& MA.PP (5%)

 

43

6330 003.550 32:: use
poo: Rom one 02:30 Rm 53:00 $8625 :02

..ma

Good Good 08.... OOQM OOQN 000;. o

_ 4 - a _ 4 a

 

 

294 968m

mags

$316.4:

Smiles

..mEOCo_

Mao

R8363: .631

6362 oz
.2354:

 

 

meazmmem Sam 23.
m ”WEDGE

1.1.

6336 09.30050 89:: ton:
poo; Ron oco 9930 Nm Escoo coczuoao =04...
.ma

 

80.8w 000.com 000.09 80.00.. 000.8 0

.1 4 _ 1 a

294 968m
mach

Asmvaasz
$863.2
Cmrcoco.

Pb

$83.3: 8m
3:63. 62

6:30:

 

 

 

waoeemfim so 3.582
m 5505

45

.695 6 804m 0388... $838: co. 006a:
6330 09.3050 30.2: .59..
683 Ron 6.6 0508 so 58:8 52.8% it

4.523th

M. N e 0

 

 

 

a 4 _ 4 1%

6.94 9.2%
mean
$891.32
Asses/42
. uoEoco_
mag
62:63 oz

 

._<_~.._E.<_2

3 031mm .5 299356
«4 “WEDGE

46

2. Discussion:

The tensile test is perhaps the most important test the
composite material must undergo, due to the test's ability to
portray the composite's overall mechanical strength and its
indication of the way the composite will perform in other
tests. How a filler affects tensile strength is dependent on
the filler's size, shape, interfacing, and packing within the
matrix (Folkes, 1980) . A very important aspect of tensile
strength is how the fibers interact with the matrix.
Additives that induce homogeneous dispersion of the fibers or
result in bonding between the phases will be apparent by an
observed increase in tensile strength and. modulus (Katz &
Milewski, 1987). The results of tensile strength and modulus
indicate MA.PP as having potential for improving the adhesion
between the recycled HDPE and wood fibers. The composite with
5% MA.PP resulted in the highest tensile strength. Elongation
at break for the composite containing 5% MA.PP is greater than
expected. This could be the result of a third phase separation
occurring from the inclusion of the additive. The composite
containing 5% MA.PP displayed a higher tensile strength than
the composite containing 2% MA.PP. However, a higher percent
elongation at break was found for the 5% MA.PP composite than
the 2% MA.PP composite. The conflicting tensile strength and
modulus data may be a result of polypropylene's separation from
polyethylene due to the two being dissimilar on a molecular

level, which causes the composite to elongate more before break

47
with increasing percent MA.PP. The inclusion of ionomer
modified polyethylene in the composite resulted in a 15%
increase in modulus and a 6.6% increase in tensile strength.
Because these results were positive, although not significant
at a .05 alpha level, ionomer"modified. PE ‘was chosen for
further study. Chlorinated polyethylene, which displayed some
positive results, was also selected for further study. The
addition of MA.PP to recycled HDPE with no fibers had no
significant effect on the resin, yet MA.PP does affect the
composite. These results provide supportive evidence for the
theory that MA.PP has the potential for improving the adhesion

between the resin and fibers.

48

cm

1. Results:

The addition of 30% wood fibers to recycled HDPE decreased
impact strength by 59% as compared to recycled HDPE alone.
Impact strength decreased for all the composites that contained
additives, as compared to the composite with no additive. The
only exception was the additive stearic acid, which exhibited
an impact strength slightly higher than the no additive
composite. (See Table 7 & Figure 5).. MA.PP's (5% and 2%)
inclusion in the composite decreased impact strength more than
any of the other additives and were significantly different.
Impact strength for 2% MA.PP was slightly lower than that of 5%

MA.PP. (See Appendix B for t-test results).

 

 

' Table 7
Izod Impact Strength
(ftlb/in)

__Material Mean SD
No Additive .269 .050
Rec. HDPE (100%) .650 .063
CPE .224 .044
Ionomer .240 .018
LDPE .226 .038
MA.PP (2%) .197 .018
MA.PP (5%) .210 .027

Stearic Acid .277 .023

 

49

633m 0242850 032:. ton...—
ooos ROM sec 023on Km 58:00 :90?on :04...
EEE
No 90 00 To We «0 rd 0

u 4 q

 

_ 4 _ 4 —

294 6:85
mean
“semis
$868.45

mag
Axoozmao: 66$
63.694 62

6:36:

 

 

 

meuznmmem SEE nos
m 5505

50

2. Discussion:

The Izod impact test determines a specimen's resistance to
breakage by flexural shock. The test measures a material's
toughness, its deformation and breaking properties. Toughness
is measured by the energy required to rupture a specimen.
Fibers will improve impact strength if they have a higher
ductility than the matrix, but most often fillers are rigid
and make the composite brittle (Clegg & Collyer, 1986). Test
results are also affected by temperature, impact velocity and
stress distribution. The relationship between filler and
matrix, and composite interfacial strength has not been
established, iHowever, one theory’ states 'that fiber-matrix
adhesion will decrease impact strength, and that impact
resistance is better for composites that have a weak interface
that will act as an energy absorbing mechanism (Clegg &
Collyer, 1986). MA.PP shows promising results for tensile
strength and modulus which might be an indication that adhesion
is occurring between the phases. If adhesion is occurring,
then the data in this case indicates that adhesion decreases

impact strength.

51

mm

1. Results:

Chlorinated polyethylene's inclusion in the composite appears
to promote water sorption in the composite material. Ionomer
modified polyethylene did not affect water sorption in any
appreciable manner. MA.PP decreased water sorption in the
composite with increasing amount of additive. The composite
that contained no additive sorbed 2.7% more water than 2% MA.PP
and 4.1% more than 5% MA.PP (based on initial weight) after
ten weeks immersed in water. Overall, the composite with no
additive gained 8.1% its initial weight in water sorbed. Five
percent and 2% MA.PP gained 4.0% and 5.4%, respectively, their
original weight after ten weeks due to water sorption. (see

Table 8 for percent moisture gain).

 

Table 8

Water Sorption
(10 weeks time)

 

 

avg. avg.
initial final

____Material w ‘ w ' h ) % Gain
NO Additive 7.78 8.4080 8.1
CPE 8.05 8.7469 8.7
Ionomer 7.87 8.4840 7.8
MA.PP (2%) 7.91 8.3360 5.4
MA.PP (5%) 8.28 8.6086 4.0

 

52

.0330 33.850 «00...... see
683 “an 96 692666 Rm 53.6... $863. .6...

 

 

 

 

AmxoBmEF
chommememmmwvwimmN—o
9‘...“
.4 .w..\
a \.
n
-. d 4 [\9 J NO
8 4 0
III - 1. ....
0.300%} .. . . .. .
4. u. .. .40
$365.: 0
Ilmll 11.1.13. 0
380:0. .. 1
m a...) .. .9 A 1 CO
v .0
who a m .
IT
95.34.02 md
mEEm

€30 9:538: .03:
Zomemmommd. £0.53

0 man—0:

53

2. Discussion:

A plastic's moisture content is closely related to its
mechanical properties, dimensional stability and appearance
(ASTM D 570, 1987). Water acts as a plasticizer for many
plastics, and tends to increase ductility and toughness, but
reduces strength and modulus. Wood fibers are highly reactive
with water, due to the large amount of hydroxyl groups present
in the structure. Chemically treating the wood fibers to
reduce their affinity for water can be done by replacing polar
hydroxyl groups with less polar groups. A water resistant
coating applied to the fibers can also help to reduce water
sorption. Generally coatings that adhere to wood are also
water sensitive (Goldstein, 1977). The hydrophylic nature of
the fibers attracts water to the interface, thus resulting in
loss of‘ mechanical properties over ‘time (Clegg & Collyer,
1986). A coupling agent could eliminate this problem.
Adhesion between the phases will reduce the amount of water
sorbed by the composite because the hydroxyl groups present on
the wood fibers are bonded and thus, will not react with the
water (Clegg & Collyer, 1986). MA.PP sorbed water, but at a
slower rate than the composite with no additive. It appears
that even if bonding is occurring with the addition of MA.PP to
the composite, that a number of the hydroxyl groups are still
free to sorb water molecules. Water sorption decreased with
increasing MA.PP content, indicating that more bonding may be

occurring with the higher percent MA.PP.

E.

1.

All of the composites exhibited creep,
fifty pounds of load.
specimens of each material were tested,
suggestive rather than conclusive.

results of the creep test.

SEEP—1&5;

Results:

54

As shown,

when subjected to

Due to a time constraint only

thus results

Figure 7 displays

the
two
are

the

the addition of additives

to the composite appears to affect the composite's creep

properties when compared to the composite with no additive.

The additives were found to reduce the extent of creep in all

 

 

of the composites in which additives were incorporated. (see
Table 9).
Table 9
Creep Analysis
Material % change (after 17 days)

No Additive

Run 1 .55

Run 2 L1; Average = .64%
CPE

Run 1 .34

Run 2 4Q1 Average = .205%
Ionomer

Run 1 .44

Run 2 $23 Average = .34%
MA.PP (2%)

Run 1 .22

Run 2 &;§ Average = .29%
MA.PP (5%)

Run 1 .13

Run 2 429 Average = .215%

 

55

6330 33.8.30 83:: .60...
c8; 5n 26 95%.; an 53:8 5838.. ..o:

 

 

 

 

 

 

AmxoovaF
o. 0.
r1 ..1 r o 0 a n e x no (a
q] q d d T d 4 d 4 q ..n o
u 4
J 80.0
.v I i .v v I 0
as V .
L .. So
Evans.
+ m UT u n. W U u .... 4 1 mac
savanna m r. .
Ilmll . « w m n. < d d u .. .
3:65. 2 N00
lol .
who l 23
Irrll .
9563.. oz 8.0
mocuc_

SE27: mmmmo
m. $505

56

2. Discussion:

The addition of fibers to the matrix reduces creep. The
percent filler will also affect creep properties. A high
content of filler in the composite will decrease the amount of
free matrix available to creep (Katz & Milewski, 1987). The
fiber-matrix interface also affects creep and the time a
composite can endure a load without breaking. Adhesion of the
fiber-matrix will allow the composite to act as one unit when
subjected to a force. A composite with a strong interface will
not pull apart as easily as one where there is no bonding of
the fibers to the matrix. The composite with 5% MA.PP creeped
less than the composite with no additive, thus adding to the
evidence that there is adhesion occurring between the HDPE and
wood fibers. But, CPE also displayed good creep results which
does not correlate to CPE's mechanical test results. Further
testing needs to be done to make an accurate interpretation.
None of the specimens failed under 'the fifty’ pound load,
although an ionomer ‘modified polyethylene sample did have

stress cracks present after three days of test.

 

57

mWWL

1. Results:

Presented in Figure 8, A-J, are the scanning electron
microscopy results for the HDPE and wood fiber composites. The
fibers appear to be relatively whole in the photos of the
various composites, and there is space between the fibers and
the matrix for all the composites with the exception of MA.PP
(5%).

Discussion:

A relatively smooth fracture surface will identify a good bond.
A fracture surface showing a number of fibers protruding from
the surface is an indication of fiber pull-out, and that stress
is not transferring to the fibers. The SEM of the MA.PP
composite with 5% additive suggested adhesion. There was
difficulty in finding any fibers protruding from the MA.PP (5%)
fracture surface. The fiber found in Figure 8-I, appears to
have pulled apart and failed, and indicates a fiber failure
rather than an adhesive failure. The space between the fibers
and the matrix was slight, and the fibers did not pull-out but
broke off at the surface. The MA.PP 2% composite SEM results
varied from that of the MA.PP 5%, and did not show signs of
adhesion. It is assumed that the material mix ratio for 2%
MA.PP was not at an optimum. Better adhesion may also result

with the 5% MA.PP if the fiber content is decreased.

SCANNING ELECTRON MI CROSCOPY
Figure 8

ii.
29k” \fiifiaaaN' 19“

 

Fig. 87A. Recycled HDPE (70%)/Wood Fiber (30%)
comp051te. 2,000 magnification.

zakU x359,
--. .e “ _J¢

 

Fig. 8—B. Recycled HDPE (70%)/Wood Fiber (30%)
composite. 350 magnification.

58

59

Figure 8 (cont.)

 

.
,
'i
J
1 I'

Fig. 8—C. Recycled HDPE (65%)/Wood Fiber (20%)[
Chlorinated PE (5%) composite. 2,000 magnification.

 

     

’ 13‘ {iv-11"“ 1:? \d;

Fig. D-D. Recycled HDPE (65%)/Wood Fiber (30%)/
Chlorinated PE (5%) composite. 350 magnification.

. ‘ -.. .1
‘ l

60

Figure 8 (cont.)

 

      

Fig. 8—E. Recycled HDPE (65%)/Wood Fiber (30%)/
Ionomer (5%) composite. 2,000 magnification.

-'-" #3,‘ I?“ -4- " . ..1. fish—- _. “I. ,/'—"' j '
“4,6126%- It). .mmsuuu— eases-
‘ , .r’ ’3‘. ~‘.. ,,,u~- ‘1. , — .
Fig 8-F. Recycled HDPE (65%)/Wood Fiber (30%)/
Ionomer (5%) composite. 350 magnification.

    

61

Figure 8 (cont.)

. N I .

iaum depb—

Fig. 8-G. Recycled HDPE (68%)/Wood Fiber (30%)/
MA.PP (2%) composite. 2,000 magnification.

 

(f). »

~ . ‘42...
1, l (. raj-1” ’ 3:7 \ .f—
4?fie£fifiigf'g.‘ . ‘ 'i
15kU _X§SB 666058
Fig. 8-H. Recycled HDPE (68%)/Wood Fiber (30%)/
MA.PP (2%) composite. 350 magnification.

 

62

Figure 8 (cont.)

 

I)» .. Aa‘ ‘ .
‘ isym unease

Fig. 8-I. Recycled HDPE (65%)/Wood Fiber (30%)/
MA.PP (5%) composite. 2,000 magnification.

    

\\ _ k -"
Q- -. ‘ 1"" '

Fig. 8-J. Recycled HDPE (65%)/Wood Fiber (30%)/
MA.PP (5%) composite. 350 magnification.

   

V. SUMMARY

SUMMARY AND CONCLUSIONS

Maleic anhydride modified polypropylene showed potential for
improving adhesion between the recycled high density
polyethylene and wood fibers. MA.PP's addition to the
composite resulted in an increase in tensile strength and
modulus, and a decrease in impact strength. Sorption of water
decreased with the inclusion of MA.PP, as did creep. Scanning
electron microscopy results also indicated the occurrance of

adhesion with 5% MA.PP.

Ionomer modified polyethylene showed some positive results but
did not appear to be inducing adhesion. Inclusion of ionomer
increased tensile strength and modulus, and had no appreciable
effect on impact strength. Ionomer' did not affect water
sorption but did decrease creep. Signs of bonding could not be

found from SEM results.

Chlorinated polyethylene gave some positive results but, like
low density polyethylene and stearic acid, it did not enhance
mechanical properties overall. Fiber pull-out and lack of

adhesion were apparent on the SEM results.

The cost of Maleic Anhydride Modified Polypropylene, or any
other additive, is an important consideration in the overall

63

64
cost of a product manufactured from a composite of recycled
HDPE and wood fibers. MA.PP's inclusion in the composite
resulted in improving mechanical properties more than the other
additives, and is thus more likely to be considered for
incorporation in the composite for end use. MA.PP has a cost
of $12.00/ lb. It is the more expensive of the five additives
analyzed (see Table 1). When using recycled materials to
construct a product, it is important to keep the cost at a
minimum, so the product can remain competitive with products

utilizing other materials, recycled or virgin.

RECOMMENDATION FOR FUTURE WORK

A replication of the research conducted for this study is
recommended in order to determine the legitimacy of the
research findings. Determination of changes in mechanical
properties after exposure to water would also be beneficial
information. A more detailed creep test with a representive
sample size in which specimen are subjected to a heavier load
is also recommended for further research. Determination of the
appropriate ratio of fiber to resin to additive so as to
achieve optimal properties is perhaps the most important area

that needs to be investigated.

65

VI . APPENDICES

APPENDIX A

MATERIAL DATA AND MANUFACTURERS

COMPOSITE CONTENTS

Recycled HDPE from master batch.

Wood Fibers--Aspen hardwood--from master batch.

Additives:

a. Chlorinated Polyethylene (CPE)--Dow Chemical Co.

b. Maleic Anhydride Modified Polypropylene (MA.PP)--
Himont Co., tradename is Hercoprime.

c. Low Density Polyethylene (LDPE)--Dow Chemical Co.

d. Stearic Acid--Sigma Chemical Co.

e. Ionomer Modified Polypropylene--Du Pont, tradename is
Surlyn.

000

Composite components: (percentage)

 

W Wood Fiber 12W
CPE 4.90 30.30 64.80
Ionomer 5.00 32.00 63.00
MA.PP 1.95 27.50 70.55
MA.PP 4.82 28.28 66.90
MA.PP 5.00 0.00 95.00
LDPE 4.85 30.09 65.06
Stearic Acid 4.95 29.79 65.26
No Additive 0.00 30.00 70.00

66

67

uanufacfurer§_2f_naferials

o Dow Chemical
2020 Dow Center
Midland, MI 48674
(517) 636-1000

o Himont USA, Inc.
1313 N. Market St.
Wilmington, DE 19894
(302) 594-5500

o Canfor Canadian Forest Products
Vancouver, British Columbia

uanufaefurers_9f_figuinment

o Baker Perkins, Inc.
901 Durham Avenue
S. Plainfield, NJ 07080
(Mnfr extruders)

o Instron Corporation
100 Royall St.
Canton, MA 02021
(Mnfr Instron)

0 Testing Machines, Inc.
400 Bayview Ave.
Amitycille, NY 11701
(Mnfr impact tester)

Du Pont

1007 Market St.
Wilmington, DE 19898
(302) 774-1000

Soltex Polymer Company
3333 Richmond Ave.
Houston, TX 77098
(713) 522-1781

Fred S. Carver, Inc.
subsid. of Sterling Inc.
W142 N9050 Fountain Blvd.
Menomonee Falls, WI 53051
(Mnfr lab press)

Polymer Machinery Corp.
154 Woodlawn Road
Berlin, CT 06037

(Mnfr lowline granulator)

APPENDIX B

DATA AND STATISTICAL ANALYSIS

STATISTICAL ANALYS IS

Students t-test (two tailed) was conducted using Epistat.
Composites containing additives were compared to the composite
with no additive to determine significant difference at a 0.05
alpha level. The samples are independent. Calculations for
the confidence limits on the difference between the means of

the samples analyzed are also given.

68

DIFFERENTIAL SCANNING CALORIMETRY DATA
Analysis of High Density Polyethylene

Percent Crystalinity

 

 

Run Material
Vir ' c c e s R
1 63.60 62.30 63.79
2 61.97 62.53 63.58
NO. 2 2 2
MEAN 62.78500 62.41500 63.68500
MED 62.78500 62.41500 63.68500
SDEV 1.152655 0.160869 0.149870
Melt Temperature (Tm)
(°c>
Run Material
Virain RecmleLJsed—Wfl
1 132.26 132.26 132.18
2 131.98 132.06 131.80
N0 2 2 2
MEAN 132.120 132.160 131.990
MED 132.120 132.160 131.990
SDEV 0.197642 0.139754 0.272431

69

DSC T-TEST RESULTS

A. Samples Compared:

Virgin(%Cryst) Recycled, Used(%Cryst)
Means 8 62.785 62.415
Variances = 1.328614 .0258789

t = .449439
df - 2

p - .6971257
The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: .3699989 +/— t(2) * 1

B. Samples Compared:

Virgin(%Cryst) Recycled, Unused(%Cryst)
Means - 62.785 63.685
Variances 8 1.328614 2.246093E-02

t 8 1.094814
df - 2

p = .3878482
The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: .9000015 +/- t(2) * 1

C. Samples Compared:

Virgin(Tm) Recycled, Used(Tm)
Means = 132.12 132.16
Variances = 3.906252E-02 1.953126E-02

t = .2263225
df = 2
p = .8419766

The MEANS of these 2 samples are NOT significantly different.
confidence limits calculation: 4.0008553-02 +/- t(2) * 1

70

71

D. Samples Compared:

Virgin(Tm) Recycled, Unused(Tm)
Means = 132.12 131.99
Variances a 3.9062528-02 7.421876E-02

t = .555859
df = 2
p = .6341908

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: .1299896 +/- t(2) * 1

TENSILE STRENGTH DATA

 

Run Material
NQ:AQQ_____QEE_____MAIRRIZi1___MAIEBI§il___IQEQE§£_
1 3574 5043.5 5565.2 4487.2 4347.8
2 3904.3 3478.3 5165.2 4874.6 4347.8
3 4530.4 3391.3 4921.7 3947.8 4608.7
4 3565.2 4899.7 6730.4 4295.7 3739.1
5 3739.1 4121.7 3913 4991.3 3565.2
6 4173.9 3700.5 4434.8 5652.2
7 4087 5426.1
8 3652.2 5043.5
9 3217.4
10 3826.1
11 4347.8
NO 6 6 11 8 5
MEAN 3914.484 4105.834 4532.801 4839.800 4121.720
MED 3821.700 3911.100 4347.800 4932.950 4347.800
SDEV 378.270 718.109 1003.911 571.402 445.953

Tensile Strength Data Cont.

Run Material
D S . Ac' e . 0
1 3405 3296.7 4782.6
2 3113.5 2838.8 5143.5
3 2838.8 2455.2 5217.4
4 3813.7 2838.8 4695.7
5 4608.7 3075.6 5095.7
6 4347.8 4912.7
7 3008.7 5095.7
8 3217.4
9
10
11
N0 5 8 7
MEAN 3555.940 3134.875 4977.615
MED 3405.000 3042.150 5043.500
SDEV 690.576 555.608 187.745

***all samples contain 5% additive and 30% wood fiber unless
otherwise specified.

72

TENSILE STRENGTH t-TEST RESULTS

A. Samples Compared:

No-Add CPE
Means = 3914.484 4105.834
Variances 8 143088.1 515670.5

t - .5774815

df = 10

p = .5763846

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 191.3501 +/- t(10) * 2.236068

B. Samples Compared:

No-Add MA.PP(2%)
Means = 3914.484 4532.801
Variances = 143088.1 1007837

t 8 1.436207

df = 15

p s .171471

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 618.3169 +/- t(15) * 3.162278

C. Samples Compared:

No-Add MA.PP(5%)
Means = 3914.484 4839.8
Variances = 143088.1 326500.6

t = 3.426167

df = 12

p - 5.020976E-03

The MEANS of these 2 samples are significantly different.
confidence limits calculation: 925.3164 +/- t(12) * 2.645751

73

74

D. Samples Compared:

No-Add Ionomer
Means - 3914.484 4121.72
Variances a 143088.1 198874

t - .8352729

df = 9

p = .4251815

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculations: 207.2363 +/- t(9) * 2

E. Samples Compared:

No-Add LDPE
Means 8 3914.484 3555.94
Variances - 143088.1 476895.2

t = 1.096798

df - 9

p = .3012076

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 358.5435 +/- t(9) * 2

F. Samples Compared:

No-Add St. Acid
Means = 3914.484 3134.875
Variances = 143088.1 308701.7

t = 2.948515

df = 12

p = 1.217783E-02

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 779.6084 +/- t(12) * 2.645751

75

G. Samples Compared:

No-Add Rec. HDPE(100%)
Means 2 3914.484 4977.615
Variances a 143088.1 35248

t = 6.582836

df = 11

p = 3.9587728E-05

The MEANS of these 2 samples are significantly different.

confidence limits calculations: 1063.131 +/- t(11) * 2.236068

TENSILE MODULUS DATA

 

(psi)
Run Material
No- mer
1 116960 208697.1 243480 118031.1 173914.3
2 165380.4 173912 240312.7 153611.4 227665.5
3 215656 173912 194784 124222.9 227665.5
4 157536.2 157668.6 252170 183851.4 243480
5 243480 187292.3 197100 185778.8 190171
6 160042.9 172120 173914.3 188817.1
7 176590.8 169565
8 170749.1 208693.3
9 208693.3
10 216426.?
11 187824
NO 6 6 11 8 5
MEAN 176509.3 178933.7 205640.5 166571.4 212579.2
MED 162711.7 173912.0 197100.0 176708.2 227665.5
SDEV 45427.8 17348.0 29156.1 32211.6 29186.2
Tensile Modulus Data Cont.
Run Material
_____LQEE_______§II_AQiQ____B§EI_ED£EllflfliI____
1 117765.7 '115122.9 115940
2 107450.7 117216 99377.1
3 122100 143370 117723.1
4 175746.? 131840 110146.7
5 204971.4 146981.8 104344
6 194785.5 109314.3
7 153040 125216
8 166960
9
10
11
NO 5 8 7
MEAN 145606.9 146164.5 111723.0
MED 122100.0 145175.9 110146.7
SDEV 42474.0 26346.6 8669.816

***all samples contain 5% additive and 30% wood fiber unless

otherwise specified.

TENSILE MODULUS t-TEST RESULTS

A. Samples Compared:

No-Add CPE
Means 8 176509.3 178933.7
Variances 8 2.063725E+09 3.009576E+08

t = .1221201

df = 10

p = .9052231

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 2424.344 +/- t(10) * 2.236068

B. Samples Compared:

No-Add MA.PP(2%)
Means 8 176509.3 205640.5
Variances 8 2.063687E+o9 8.500776E+08

t 8 1.620503

df 8 15

p - .1259516

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculations: 29131.13 +/- t(15) * 2.23607

C. Samples Compared:

No-Add MA.PP(5%)
Means = 176509.3 166571.4
Variances = 2.063687E+09 1.037587E+09

t 8 .4807461

df = 12

p = .6393428

The MEANS of these 2 samples are NOT significantly different.
confidence limits calculation: 9937.953 +/- t(12) * 2.236068

77

78

D. Samples Compared:

No Add Ionomer
Means 8 176509.3 212579.2
Variances = 2.063687E+09 8.518331E+08

t 8 1.525324

df 8 9

p = .1615186

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 36069.88 +/- t(9) * 2.236068

E. Samples Compared:

No Add LDPE
Means 8 176509.3 145606.9
Variances = 2.0637ZSE+09 1.804041E+09

t 8 1.156193

df 8 9

p = .2773704

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 30902.42 +/- t(9) * 2.236068

F. Samples Compared:

No Add St. Acid
Means 8 176509.3 146164.5
Variances = 2.063725E+09 6.941386E+08

t = 1.579911

df 8 12

p = .1401119

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 30344.83 +/- t(12) * 2.236068

79
G. Samples Compared:
No Add

Means 8 176509.3
Variances 8 2.063725E+09

t 8 3.721657
df 8 11
p 8 3.371885E-03

Rec. HDPE(100%)

111723
7.516706E+07

The MEANS of these 2 samples are significantly different.

confidence limits calculation:

64786.31 +/- t(11) * 2.236068

PERCENT ELONGATION DATA

 

Run Material
Nam—MWL—MALW
1 1.7 3 2 4.5 3.5
2 1.6 3 1.38 4.375 2.25
3 1.7 2.4 2.38 4 3.13
4 1.5 4.25 2.82 3 4
5 1.1 2.5 4.63 2.875 2.63
6 .8 3.13 3.25 3.875
7 3 4.125
8 2.75 3.25
9 1.75
10 2.25
11 3.75
NO 6 6 11 8 5
MEAN 1.400000 3.046667 2.723637 3.750000 3.102000
MED 1.550000 3.000000 2.750000 3.937500 3.130000
SDEV 0.368783 0.659749 0.931153 0.626783 0.691787
Percent Elongation Data Cont.
‘ Run Material
_____JJEEL_____iEa_AEiQ___B§£1_Hflfifiilflflil_____
1 4.5 4 308.75
2 4.25 4.125 142.5
3 3.25 2.75 246.25
4 2.875 3.625 263.75
5 3.875 4.5
6 5.25
7 4.125
8 4.25
9
10
11
NO 5 8 4
MEAN 3.750000 4.078125 240.313
MED 3.875000 4.125000 255.000
SDEV 0.678924 0.713200 70.32200

***all samples contain 5% additive and 30% wood fiber unless
otherwise stated.

PERCENT ELONGATION t-TEST RESULTS

A. Samples Compared:

NO Add CPE
Means 8 1.4 3.046667
Variances 8 .1360005 .4352692

t8 5.336556

df 8 10

p 8 3.299713E-04

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 1.646667 +/- t(10) * 2.236068

B. Samples Compared:

No Add MA.PP(2%)
Means 8 1.4 2.723637
Variances 8 .1360005 .8670456

t 8 3.30328

df 8 15

p 8 4.82626E-03

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 1.323637 +/- t(15) * 3.162278

C. Samples Compared:

No Add MA.PP(5%)
Means = 1.4 3.75
Variances = .1360005 .3923572

t 8 8.13894

df 8 12

p 8 3.099442E-06

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 2.350001 +/- t(12) * 2.645751

81

82

D. Samples Compared:

No Add Ionomer
Means 8 1.4 3.102
Variances 8 .1360005 .478569

t 8 5.235242

df 8 9

p - 5.3819llE-04

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 1.702 +/- t(9) * 2

E. Samples Compared:

NO Add LDPE
Means 8 1.4 3.75
Variances 8 .1360005 .4609375

t8 7.328751

df 8 9

p 8 4.421663E-05

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 2.350001 +/- t(9) * 2

F. Samples Compared:

No Add St. Acid
Means 8 1.4 4.078125
Variances = .1360005 .5086496

t 8 8.341909

df 8 12

p 8 2.503395E-06

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 2.678126 +/- t(12) * 2.645751

83

G. Samples Compared:

No Add Rec. HDPE(100%)
Means 8 1.4 240.3125
Variances 8 .1360005 4945.184

t 8 8.594624

df 8 8

p 8 2.598763E-05

The MEANS of these 2 samples are significantly different.

confidence limits calculation: 238.9125 +/- t(8) * 1.732051

é"

._____Bg2L_flQ2EL1QQiL_______Bg2L-flDEE-i-fiiuALE£______

MA.PP + HDPE

Tensile Strength

Material

1 4782.6 4347.8

2 5043.5 4817.4

3 5217.4 4521.7

4 4695.7 4695.7

5 5095.7 4782.6

6 4921.7 4991.3

7 5095.7 4469.6

8 5217.4
N0 7 8
MEAN 4978.900 4730.436
MED 5043.500 4739.150
SDEV 187.247 286.532

Tensile Modulus
Run Material
0 . +

1 115940 112375.4

2 99377.1 113832.7

3 117723.1 187824

4 110146.7 113832.7

5 104344 118264

6 109314.3 147830

7 125216 123644.4

8 120160
NO 7 8
MEAN 111723.0 129720.4
MED 110146.7 119212.0
SDEV 8669.816 26102.3

84

85

Izod Impact

Run Material
MW—
1 .655 .445
2 .707 .292
3 .624 .275
4 .648 .363
5 .704 .305
6 .527 .423
7 .72 .322
8 .615 .282
No 8 8
MEAN 0.650000 0.338375
MED 0.651500 0.313500

SDEV

0.063386

0.065317

MA.PP + REC. HDPE t-TEST RESULTS

A. Samples Compared:

Rec. HDPE (TS) MA.PP + Rec. HDPE (TS)
Means 8 4978.9 4730.436
Variances 8 35061.33 82100.6

t 8 1.953568

df 8 13

p = 7.261483E-02

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 248.4639 +/- t(13) * 2.645751

B. Samples Compared:

Rec. HDPE (TM) MA.PP + Rec. HDPE (TM)
Means 8 111723 129720.4
Variances 8 7.516571E+07 6.813285E+08

t 8 1.735326

df 8 13

p = .1063097

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation: 17997.35 +/- y(13) * 2.645751

C. Samples Compared:

Rec. HDPE(100%)(I) Rec. HDPE + 5%MA.PP(I)
Means 8 .65 .338375
Variances 8 4.017796E-03 4.266245E-03

t 8 9.684035

df 8 14

p 8 < 10 (86)

The MEANS of these 2 samples are significantly different.

confidence limits calculation: .311625 +/- t(14) * 2.645751

86

IZOD IMPACT DATA

 

(ftlb/in)
Run Material
MAM—MWW
1 .246 .213 .174 .24 .243
2 .352 .31 .175 .22 .21
3 .199 .184 .194 .197 .256
4 .294 .22 .198 .24 .225
5 .318 .214 .206 .167 .256
6 .262 .205 .213 .205 .248
7 .23 .22 .184
8 .251 .23
N0 8 6 7 8 6
MEAN 0.269000 0.224333 0.197143 0.210375 0.239667
MED 0.256500 0.213500 0.198000 0.212500 0.245500
SDEV 0.049624 0.043803 0.017743 0.026774 0.018490
Izod Impact Data Cont.
Run Material
______LDBE_____5E1_AQiQ___B§£1_HD£EIIQQ&1_____
1 .221 .318 .655
2 .166 .262 .707
3 .246 .256 .624
4 .196 .252 .648
5 .271 .275 .704
6 .194 .285 .527
7 .253 .29 .72
8 .262 .615
NO 8 7 8
MEAN 0.226125 0.276857 0.650000
MED 0.233500 0.275000 0.651500
SDEV 0.037794 0.023126 0.063386

***all samples contain 5% additive and 30% wood fiber unless

otherwise stated.

87

IZOD IMPACT t-TEST RESULTS

A. Samples Compared:

NO Add CPE
Means 8 .269 .2243333
Variances 8 2.462575E-03 1.918674E-03

t 8 1.749076

df 8 12

p - .1057843

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation:4.466667E-02 +/- t(12) * 2.6458

B. Samples Compared:

NO Add MA.PP(2%)
Means 8 .269 .1971429
Variances 8 2.462575E-03 3.14802E-04

t8 3.619657

df 8 13

p 8 3.112641E-03

The MEANS of these 2 samples are significantly different.

confidence limits calculation:7.185714E-02 +/- t(13) * 2.6458

C. Samples Compared:

NO Add MA.PP(5%)
Means 8 .269 .210375
Variances 8 2.462575E-03 7.168397E-04

t 8 2.940719

df 8 14

p = 1.074117E-02

The MEANS of these 2 samples are significantly different.

confidence limits calculation: .058625 +/- t(14) * 2.645751

88

89

D. Samples Compared:

No Add Ionomer
Means 8 .269 .239667
Variances 8 2.462583E-03 3.418746E-04

t 8 1.366889

df 8 12

p - .196722

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation:2.033334E-02 +/- t(12) * 2.6458

E. Samples Compared:

No Add LDPE
Means 8 .269 .226125
Variances 8 2.462575E-03 1.428408E-O3

t 8 1.944101

df 8 14

p = 7.225883E-02

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation:4.287499E-02 +/- t(14) * 2.6458

F. Samples Compared:

No Add St. Acid
Means 8 .269 .2768572
Variances 8 2.462575E-03 5.348225E-04

t 8 .3827987

df 8 13

p = .7080548

The MEANS of these 2 samples are NOT significantly different.

confidence limits calculation:7.857144E-03 +/- t(13) * 2.6458

90

G. Samples Compared:

No Add Rec. HDPE(100%)
Means 8 .269 .65
Variances 8 2.462575E-03 4.017796E-03

t 8 13.38662

df 8 14

p 8 < 10 (-5)

The MEANS of these 2 samples are significantly different.

confidence limits calculation: .381 +/- t(14) * 2.645751

Time (days)

WATER ABSORPTION DATA

(Average Moisture Gain in grams)

Material

NO Addi;i!g____QEE_____HALEELZil___HA&EEL§il———IQDQEEI

14

21

28

35

42

49

56

70

0.0950

0.1330

0.1650

0.2232

0.3094

0.3890

0.4448

0.4836

0.5230

0.5625

0.5860

0.6280

0.0413

0.1273

0.2173

0.2798

0.3788

0.4663

0.5263

0.5838

0.6213

0.6538

0.6769

0.6969

91

0.0200

0.0700

0.1070

0.1330

0.1800

0.2200

0.2500

0.2810

0.3170

0.3480

0.3790

0.4260

0.0100

0.0270

0.0550

0.0790

0.1130

0.1510

0.1790

0.2046

0.2316

0.2636

0.2906

0.3286

0.0570

0.1030

0.1820

0.2360

0.3120

0.3820

0.4350

0.4850

0.5120

0.5490

0.5800

0.6140

CREEP ANALYSIS DATA

(gain in inches)

Time(days) Material

New—MWW

 

.021 0.0000 0.0000 0.0150 0.0015 0.0025

.083 0.0045 0.0065 0.0150 0.0015 0.0095
1 0.0130 0.0075 0.0185 0.0025 0.0145
2 0.0130 0.0075 0.0185 0.0025 0.0155
3 0.0130 0.0075 0.0185 0.0025 0.0155
4 0.0130 0.0075 0.0195 0.0050 0.0155
5 0.0155 0.0075 0.0195 0.0070 0.0155
6 0.0155 0.0075 0.0195 0.0100 0.0155
7 0.0180 0.0075 0.0195 0.0100 0.0155
14 0.0255 0.0090 0.0195 0.0100 0.0155
17 0.0255 0.0090 0.0195 0.0100 0.0155

92

VII . BIBLIOGRAPHY

LIST OF REFERENCES

Agarwal, Bhagwan D., and Iawerence J. Broutman. Anelye1e_eng
' Toronto: John Wiley and

Sons, 1980.

"Analyst: Solid Waste Becomes Crisis." MQQQIE__Bla§§iQ§1
(January, 1988): 25-26.

"ASTM D 570, Water Absorption of Plastics." W
AfiIu_§§andaId§1 8-01 (1987): 187-190~

Brennan, W. P. "Characterization of Polyethylene Films by

Differential Scanning Calorimetry."

Appliee;ien_§§gdy_zge Norwalk, Ct.: Perkin-Elmer Corporation,

Instrument Division, March, 1978.

Clegg. D-W. and A-A- Collyer, ueghanigal__£renerfies__2f
Wasting. London: Elsevier Applied Science
Publishers, 1986.

Dalvag, H., C. Klason and H. E. Stromvall. "The Efficiency of
Cellulose Filler in Common Thermoplastics Part II: Filling With

Processing Aids and Coupling Agents." Ingernegienelelegzne1_efi
Eglxmeri__uaferial§1 11 91985): 9- 38.

Folkes. M-J- 3h2rf__Iiber__Beinferged__lnermenla§tigsi
Chichester: Research Studies press, 1982.

Goldstein, Irving 8., ed. °
Washington D.C.: American Chemical Society, 1977.

Hua, Li and Per Flodin. "Cellulose Fiber-Polyester Composites
With Reduced Water Sensitivity(2)--Surface Analysis." Eelyme;

92322515251 8 (1987): 203-207-

Hua, Li, Pawel Zadorecki and Per Flodin. "Cellulose Fiber-
Polyester Composites with Reduced Water Sensitivity ( 1) --
Chemical Treatment and Mechanical Properties . " flame:

sensesitesi 8 (1987): 199-202-
Hull, Derek. A d ct'o o ' ' .
Cambridge: Cambridge University Press, 1981.
Katz, Harry S. and John V. Milewski. andb o F s
s or s ' New York: Van Nostrad Reinhold

Company, 1987.

93

94

Klason, C., J. Kubat. and. H.E. Stronvall. "Efficiency of
Cellulose Filler in Common Thermoplastics Part I: Filling
Without Processing Aids or Coupling Agents" . We;

I2nrnal_of_£olxmeric_uaterial_1 11 (1984)=159-187

Kokta, Bohuslav, Claude Deneault and Alphons D. Beshay. "Use
of Grafted Aspen Fibers in Thermoplastic Composites: IV Effects
of Extreme Conditions on. Mechanical Properties." Eelyme;

compositesi 7 (1986): 337-348-

Mark Herman, Norbert Bikales, Charles Overberger and Gorg
Menges. "Maleic Anhydride." ‘

eng_Engineeringe 2nd ed. New York: John Wiley and Sons, 1981.

Melosi, Martin V. "The Cleaning of America." Enyizenmenge 23
(1981): 6-13, 41-43.

Michell, Anthony J., Janet E. Vaugham and Donald Willis. "Wood
Fiber--Synthetic Polymer Composites. I. Laminates of Paper and
Polyethylene." o e ce: o ' o 5
(1976): 143-154.

Michell, Anthony J., Janet E. Vaugham and Donald Willis. "Wood
Fiber-Synthetic Polymer Composites. II Laminates of Treated

Fibers and Polyclefine-" I22rna1_of_Applied_£21¥mer_§ciencei
22 (1973): 2047-2061.

"Milk Bottles Reembodied." £eekeg1ng_pigee§e September, 1987:
82 & 84.

Pattanakul, Chahe. "Characteristics Changes In Recycled HDPE
From Milk Bottles." Diss. Michigan State University, 1987.

Pineri, Michel and Adi Eisenberg, ed. Stru t e ' s
efi_lenemeree Dordrecht, Holland: D Reidel, 1986.

Sanschagrin, B., S. T. Sean and B. V. Kokta. "Mechanical
Properties of Cellulose Fibers Reinforced Thermoplastics."
W The Society of

Plastics Industry, Inc., 1988.

Schneidman, Diane. "Plastics: Progress and Peril." Mezketing
Henge 21 (1981): 6-8.

Serie, Terry L. and Ann H. Mattheis. Letter to State
Government Relations Committee. October 6, 1988.

Wright, Timothy L. c e ' s 8 'c ' d
Dexeloementi 1987.
Zadorecki , Pawel and Per Flodin . " Properties of Cellulose-

Polyester Composites." golymer Qomposites. 7 (1986): 170-175.

GENERAL REFERENCES

"ASTM D 256-81, Impact Resistance of Plastics and Electrical

Insulating Materia18-" W 8.01
(1987): 81-102.

"ASTM D 2990-77, Tensile, Compression, and Flexural Creep and

Creep-Resistance of Plastics." Annnal_Beok_of_A§TM_§tandard§1
8.02 (1988): 516-523.

"ASTM D 638-77a, Tensile Properties of Plastics." Anngel_§eek
9f_ASTM_§tandard§1 8-01 (1987): 227-238-

Auchter, Richard J. "Trends and Prospects for Use in Fiber

Products." Asnen1.3xmnosium_2roceedingei 1972.

Beshay, Alphons D. , Bohuslav V. Kokta and Claude Deneault.
"Use of Wood Fibers in Thermoplastic Composites II:

Polyethylene." BEIYEQI.QQEEQ§1§§§ 6 (1985): 261-271.
Bikales, Norbert M., ed. AW New York:

Wiley-Interscience, 1971.

Broutman, Lawrence J., and. Richard. H. Krock, eds juggezn

Qempeei;e_fle§e;ielee Reading, Mass.: Addison-Wesley Publishing
Company, 1967.

Crawford, R.J. Eleegiee__zngineeringe 3rd ed. New York:
Pergamon Press, 1987.

39.151354. Computer Software. Tandy Corp., 1983-1986. Tandy
version 03.20.01., disk.

Fettes, Iii-M., ed. W New York:
Interscience, 1964.

Garratt, George A. The Meehenical Propezti es 9; Weoe. New

York: John Wiley and Sons, Inc., 1931.

Graham, Samuel Alexander, Robert P. Harrison Jr., Casey E.
Westell Jr. We; Ann Arbor, MI: University of Michigan
Press, 1963.

Kenig, S. "Fiber Orientation development in Molding of Polymer
Composites.“ Polymer Composites. 7 (1986): 50-55.

Kennedy, J.F., G.O. Phillips, D.J. Wedlock and P.A. Williams,
ed. Qellulose and its Derivetives: Chemistry. Bioehemietry ene

95

96
Applieefiienee Chichester: Ellis Horwood Limited, 1985.

Leaversuch, Robert D. "Industry Begins To Face Up To The
Crisis Of Recycling." MW (March, 1987): 44- 47.

Quynn, Richarrd G. "Fiber Composites Get Set To Take Off."
31W (September/October 1986): 50-57.

Ramirez, Amando Padilla and Antonio Sanchez Solis.
"Development of a New Composite Material From Waste Polymers,

Natural Fibers and Mineral Fillers." W
W3. 29 (1984): 2405- -2412-

Roumpf, Judy. "The Plight of Plastic Recycling." W
W (1984): 12, 13 & 31.

Sliker, Alan PhD. Personal interview. 8 November 1988.

Tratina, Gerald G. "Creep Analysis of polymer Structures."
MW 26 (June. 1986): 776-780.

Worthy, Ward. "Wide Variety of Applications Spark Polymer
Composite Growth.” Q g EN: (March 16, 1987): 7-13.

"@0000?