. . . . . . . . V . . . ..... .........
.... . ... . . . . .1 .. .. .... v
.. .. . . .. . . . . . .. l... .... .
. . . . . ..
. . . .. .
. . I . . . .. . . .
. . _ . . . . . ..
.. . .. .. . .. . . .. . ..
. ... .. . . . .
. _ . . : . . .. ..
Q . .. . .
. .. .. . . . . . . .
. I ... .. . . . V . . . . ..
.. .... . .. .... . . . . . .
. . . . . . . v
v . .. .. . . . . . . . .
. .. . ~ . . . . .. .
. . . . . . . . . . n
._ . .
. . . . . . . . . .
. . . . . . . ,. .
. . . _ . . . . .. ,. .
V .. .. . . .. .
.. ... . . ., . .
. . .. .
u . . . . . . .
. .. .. . . . . . . .
. . .. ... .. . . . .
. . . .. .
. . .
. .. . . ... . .
. . .. . .

 

 

 

..”.-pp“"-n11u~vr-

 
 
 
  

 

 

. lw I,t>.l ' flirt I

 

*urgtg

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[II I“ llllllilllllllllflllllll [IUINHIIHHI
3 1293 00567 8838
LIBRARY

1 17 I 70 M
Michigan State
University

 

 

 

 

This is to certify that the

thesis entitled

TV . .--- -- flan-C T “ Mr PT"p‘C-TTF"V 7“?“ Ah“. 1"”?ch
I- k-(xJ- v.1. (1-. k. -t:l_ i-J(/'Lvt ,'.- _Lk,‘ ,..' L ;. .‘. 4.....JJ..n.. ‘ko'.lO-L-A-
fi,-\1--- {WI-3‘ ' ~ ‘. -»_Vv'\ ’Tn j-v (YN T—‘~«-v-‘ .I... (I
rlxln Jk.“A .1JL'IIZ..JI.-\I naCKxJ/ .h‘ 1-8.1.4-.53
1"]:31 =T
" x" (1'37"! T\_ ***** 1 (j I *‘f’w {‘1'“ ‘ " (‘,"1f‘j':“lt“‘ :‘-:rrl]Tj) L‘fip‘plfil
.;k..."./ f. QIJLJ. f i .UP‘J LIV .5 [J .L .. I. c-k.) \JJ.\.A- . .1 4x4. .;~ 4.)..4 ‘.

presented by
Linoy lunar Goaoi

has been accepted towards fulfillment
of the requirements for

IRS. . degree mm”? ~ “T

 

7'44... 51/4

Major professor

Susan E. Selke

 

Date April 14, 1989

0.7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

MSU

 

 

 

‘ RETURNING MATERIAIS:
Place in book drop to remove this
checkout from your record. FWES

 

LIBRARIES will be charged if book is returned
=- after the date stamped below.
M 1993: D . . , .
' I'LD Q 3.! MAY 2 4. 2000
1 67 . ‘7 _ f
“4311‘ 2.3:. risgfizfl“) ‘
x t! ‘2? MAG 2
if." 33 43 DE 2 1 1998
#5- 33 B I “ :Ev‘q
T (M? 3153: 0,6! Iii 93:9
rm 1 6 1996 'MY 2 4.2000
w ,‘skgggm
U .

 

 

 

PROCESSING-MORPHOLOGY-PROPERTY RELATIONSHIPS
FOR COMPOUNDIN G WOOD FIBERS
WITH
RECYCLED HDPE USING A TWIN-SCREW EXTRUDER

By

Binoy Kumar Gogoi

A THESIS

submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

School of Packaging
1989

ABSTRACT

PROCESSING-MORPHOLOGY—PROPERTY RELATIONSHIPS
FOR COMPOUNDING WOOD FIBERS
WITH
RECYCLED HDPE USING A TWIN-SCREW EXT RUDER.

By

Binoy Kumar Gogoi

The development of composites consisting of wood fibers
and recycled plastics offers not only an opportunity to utilize
an abundant natural resource but also a means to alleviate the
serious problem of plastics waste disposal. In this.study,
aspen fibers are incorporated into recycled high density
polyethylene using a co-rotating intermeshing twin-screw
extruder. Mixing conditions are varied to study the effects of
fiber pre-treatment, screw configuration and compounding
temperature on the mechanical properties of the composite.
Tensile, impact and flexural strength are measured as a
function of fiber concentration while the fracture surfaces of
tensile specimens are studied using the SEM. Theoretical
predictions for tensile strength and tensile modulus are

compared with experimental data.

To my father and mother, Jogneshwar and Bijoya Gogoi.

iii

ACKNOWLEDGEMENTS

I would like to offer my sincere gratitude to Dr. Susan
Selke, PhD. (School of Packaging, MSU), my major adviser, and
my committee members, Dr. Paul Singh, PhD. (School of
Packaging, MSU) and Dr. Dahsin Liu, PhD. (Department of
Metallurgy Mechanics & Material Science, MSU) for their
guidance and academic support. I also thank Dr. Kit Yam, PhD.
(Department of Food Science, Rutgers University) for his

suggestions and cooperation.

B.K.Gogoi
5.10.1989

iv

TABLE OF CONTENTS

P a 8 e
LIST OF TABLES .
LIST OF FIGURES

I. INTRODUCTION
II. SHORT FIBER REINFORCED PLASTICS .
Ill. COMPOUNDING TECHNIQUES
A. GENERAL INFORMATION ON COMPOUNDING
1. Continuous versus batch process
2. Equipment available :
a). Banbury Mixer
b). Continuous Mixer
c). Two Roll Mill .
d). Kneader
B. MELTING MECHANISM
C. EXTRUSION FOR FIBER REINFORCEMENT .

viii

11
11
12
12
12
14
14
14
17

1. Single Screw versus Twin-Screw Extruder18

2. Twin screw extruder.

D. FEATURES OF EXT RUSION COMPOUNDING

19
23

IV. MATERIALS AND METHODS
A. MATERIALS
1. Polymer matrix
2. Filler .
B. METHODS .
1. Processing
2. Properties
a). Tensile properties .
b). Impact strength
0). Flexural properties.
(1). Fiber length distribution .
e). Fracture surface
V. RESULTS AND DISCUSSIONS
A. EFFECT OF FIBER TREATMENT
1. Tensile strength, tensile modulus,
and elongation
2. Impact strength
3. Flexural properties
4. Selection of fiber from pre-treatments
B. EFFECT OF SCREW CONFIGURATION
C. FIBER LENGTH IN TERMS OF MOMENTS
OF DISTRIBUTION .
D. EFFECT OF COMPOUNDING TEMPERATURE
1. Tensile strength and tensile modulus
2. Flexural strength and flexural modulus

E. SCANNING ELECTRON MICROSCOPY

vi

24
24
24
24
24
26
27
28
28
29
29

30

30
36
36
39
39

49
49
49

.51

51

F. THEORETICAL PREDICTIONS VS. EXPERIMENTAL DATA 6 0

1. Tensile Strength . . . . 6 0
2. Tensile Modulus . . . . 6 7

VI. SUMMARY
A. SUMMARY AND CONCLUSIONS . . . 7 5
B. SUGGESTIONS FOR FURTHER WORK . . 7 6

VII. APPENDIX
A GLOSSARY . . . . . 7 8

VIII.BIBLIOGRAPHY . . . . . 8 0

vii

LIST OF TABLES

 

 

Table Item Page
I I I
1. Details of screw configuration . . 31
2. Average fiber length and polydispersity 50

3. Coeff. of correlation for heat treated aspen

fibers(Tensile strength) . . 63
4. Coeff. for heat treated aspen fibers

(Tensile strength) . . . . 64
5. Coeff. for heat treated aspen fibers

(Tensile modulus) . . . . 72
6. Coeff. for acetylated aspen fibers

(Tensile modulus) . . . . 73
7. Coeff. for untreated aspen fibers

(Tensile modulus) . . . . 74

 

viii

LIST OF FIGURES

 

 

Figure Item Page
I I I I
1. Schematic view of an extruder 13
2. Melting mechanism of a polymer in a single-screw
extruder 15
3. Co-rotating twin-screw operation 20
4. Counter-rotating twin-screw operation 20
5. Flow profile for twin-screw extruder 21
6. Rotary Grinder :: setting arrangement of blades 25
7. Tensile strength vs fiber content(SC 1, 150 °C) 32
8. Tensile modulus vs fiber content(SC 1, 150 °C) 33
9. Elongation at break vs fiber content(SC 1, 150 0C) 34
10. Izod impact strength vs fiber content(SC 1, 150 0C) 37
11. Flexural yield strengh vs fiber content(SC 1, 150 °C) 38
12. Flexural modulus vs fiber content(SC 1, 150 0C) 40
13. Agitator design worksheet 41
14. Probability density curves for fiber length 43
15. Effect of SC on tensile strength(untreated fiber, 150°C)45
16. Effect of SC on tensile modulus(untreated fiber, 150°C)46
17. Effect of SC on flexural yield strength

(untreated fiber,150°C)

ix

47

18.
19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

Effect of SC on flexural modulus (untreated fiber, 150 0C) 48

Effect of compounding temp. on tensile strength
(untreated fiber, SC 1)

Effect of compounding temp. on tensile modulus
(untreated fiber, SC 1)

Effect of compounding temp. on flexural yield
strength (untreated fiber, SC 3)

Effect of compounding temp. on flexural modulus
(untreated fiber, SC 3)

SEM :: 32 % untreated fibers,150 °C, SC 4

SEM :: 30 % untreated fibers,150 oC, SC 1

SEM :: 25 % untreated fibers,150 °C, SC 1

SEM :: 25 % untreated fibers, 210 0C, SC 1

SEM :: 38.5 % heat treated fibers,150 oC, SC 1 .
SEM :: 28 % acetylated fibers,150 °C, SC 1
Correlation of theory with tensile strength data
(Heat treated aspen fibers in HDPE matrix)
Correlation of theory with tensile strength data
(Untreated aspen fibers in HDPE matrix)
Correlation of theory with tensile strength data
(Acetylated aspen fibers in HDPE matrix)
Correlation of theory with tensile modulus data
(Heat treated aspen fibers in HDPE matrix)
Correlation of theory with tensile modulus data
(Acetylated aspen fibers in HDPE matrix)
Correlation of theory with tensile modulus data

(Untreated aspen fibers in HDPE matrix)

X

52

53

54

55

57

57

58

58

59

59

62

65

66

69

70

71

INTRODUCTION

INTRODUCTION

In recent years, polymer composites containing wood
fibers have received considerable attention both in the
literature and from industry (Belmaris et al., 1981; Pal et a1,
1985; Semsarzadeh, 1985, 1986; Ramirez and Solis, 1984;
Rowlands et al., 1986). This is because wood fibers are strong,
light weight, non-abrasive, non-hazardous, inexpensive (Majali
et al., 1979) and can serve as an excellent reinforcer or
extender for plastics (Cruz - Ramos, 1986). Wood fibers are also
plentiful and renewable; in the United States, one third of the
land is covered by forest ( Youngquist, 1983). Wood fibers, as
fillers or reinforcing agents, can alleviate the rising cost and
possible shortage of plastics (W 1986) and can
help to satisfy the increasing demand for inexpensive high-
performance building materials (Sheldon, 1982; Semsarzadeh,
1985). Wood fiber composites can be prepared by extrusion
and compression/injection molding to produce a variety of
products which may be used in packaging, paper products,
building materials, automobile parts, etc. (Soule and
Hendrikson, 1966; Kokta et.al., 1986). The use of wood fibers is
relatively limited (Pal et. al., 1984) although talc, mica, clay,
starch, rice husks an wood flour are commonly used as fillers
for plastics (Dahl, 1948; Shepherd et al., 1974; Westhoff, 1974;
Otey, 1976; Lightsey, 1981; Theberger, 1981; Willis and King,
1981). Combined with plastics, the short hardwood fibers have

the potential for developing a new type of composite _ one
1

2

that could offer good strength and mechanical properties, and
ease of moldability.

Increased emphasis has also been placed on the recycling
of plastic wastes because of the highly publicized problems
associated with waste disposal. The availability of landfill space
has decreased rapidly and the cost of land-filling plastic wastes
has increased enormously while the "NIMBY" (Not In My Back
Yard) phobia rises in all its severity to reject any potential site
for waste disposal (Selke, 1987). Environmentalists, journalists,
industrialists, researchers and others are showing concern over
the threat of water and air pollution as US citizens continue to
generate more than 410,000 tons of solid waste every day.
Plastic packaging, which constitutes 4 % to 7 % (6 million tons
per year) of this consumer generated waste, has become a
major target of public attack because of its high visibility
(Packaging Technology, 1988). Most of this six million plastic
packaging waste consists of items like HDPE milk bottles, PET
bottles and polystyrene packages which have short use cycles.
Generally post-consumer HDPE has properties similar to those
of virgin but is sold for less than half the price (Yam, 1987). In
terms of economic value, considering the average value of
recycled plastics as $0.20 per pound, the amount of waste is

$2.4 billion a staggering amount which calls for proper

investigation and exploitation. The National Academy of
Sciences estimates that 52 million pounds of packaging
materials are being dumped into the ocean each year and these

together with other plastic solid wastes are responsible for the

3

yearly death of one to two million sea birds, 100,000 sea
mammals and countless fish (Packaging Technology, 1988).

The increasing concerns for the disposal of plastic waste
have led to recycling as a priority in most waste management
programs. Large quantities of recycled plastics may be
available as feedstock in the future and HDPE milk bottles show
considerable promise in this regard. In 1987, 740 million
pounds of blow-molded HDPE milk bottles along with another
1.357 billion pounds of other HDPE blow-molded bottles were
produced in the United States (Modern Plastics, 1988). The
disadvantage of recycled plastics is that they are of lower
grade due to contamination and degradation, and thus they are
not suitable for applications related to food products.
Nevertheless recycled plastics can be used as an effective
matrix for polymer composites, such as those containing wood
fibers. Woodhams et al. (1984) has shown that high density
polyethylene (HDPE) reinforced by thermomechanical pulp
(TMP) has a stiffness/weight ratio equal to or exceeding those
of steel, aluminum and fiberglass composites. We also find that
composites containing wood fibers are already finding
applications as pipes, channels and housing. Short hardwood
fibers are preferred because they are thicker walled and less
likely to break during compounding.

The objective of this research is, therefore, to investigate
the processing - morphology - property relationships for
compounding wood fibers with recycled HDPE using a twin-

screw extruder. It is necessary to study. the effects of

4

processing conditions on the mechanical properties of wood
fiber-recycled HDPE composites so that it may be possible to
identify the optimum processing conditions which will give an
end product with the best properties.

The present understanding of fabrication techniques,
bonding mechanisms, mechanical properties, effect of moisture
and weathering resistance etc., for wood fiber reinforced
plastics is incomplete and scattered (Landel, 1981; Youngquist,
1983; Marra, 1984). Particularly, very little has been studied
on the effects of processing history on recycled plastics. A
twin-screw extruder with two stages can be used to effectively
incorporate the wood fibers into polymers : the polymer is
melted in the first stage and the fibers gently blended in the
second stage to minimize fiber breakage. Good mechanical
properties of short-fiber-reinforced thermoplastic composites
depend on the extent of mixing of the fibers and the polymer
matrix. While preparing fiber reinforced composites using
extrusion techniques, it is necessary to optimize the amount of
compounding : inadequate compounding causes poor dispersion
and poor wetting of the fibers while excessive compounding
causes severe fiber damage _ both leading to reduced
mechanical strength.

This thesis has been organised in ten main sections. The
introduction is followed by a brief review of the theoretical
concepts governing short fiber reinforced plastics. The third
chapter outlines the prevalent techniques of compounding and

brings out the essential features of a twin-screw extruder

5

which makes it specifically suitable for compounding shear
sensitive fibers. Materials & Methods are discussed in chapter'
IV. The experimental results presented in Chapter V first
compare the mechanical properties of aspen fibers with three

different types of treatment untreated, heat treated and

acetylated. _ and then evaluate the fiber with the best
properties and least cost. Theoretical predictions are correlated
with experimental results in this chapter. Conclusions and
suggestions for further work are on pages 75, 76 and 77. A

glossary of important terms is included as an appendix.

II . SHORT FIBER REINFORCED PLASTICS

7

At the critical aspect ratio, both the fiber and the matrix
will fracture at the same plane. Additional increase in the fiber
length will not increase the strength of the composite as
failure will take place at the critical aspect ratio. If the fibers
are shorter than the critical aspect ratio, full load transfer will
not be achieved and there will be fiber pull out during tensile
stress. The polymer will slide past these short fibers instead of
breaking them until all the ductility is lost and the polymer
fractures. Inadequate wetting of the fibers by the polymer
matrix creates weak sites at the polymer ends, causing
reduction in fiber-matrix reinforcements. To achieve maximum
reinforcing efficiency, the fibers must be at least 10 times
longer than the critical length (Cloud, 1975; Wall, 1987).
Bonding the filler to the resin by means of a coupling agent
helps to prevent phase separation and results in a stronger
material (Beshay et. al., 1985). Studies on the effect of wood
fibers incorporated into cement using coupling agents are
reported by Gautts and Campbell (1979). The results of the
investigation confirm that the mechanical performance of wood
fiber reinforced cement composites can be altered by the use of
coupling agents. Promising mechanical properties have also
been cited for composites of grafted olefin polymers and
cellulose fibers (Coran and Patel, 1982; Goettler, 1983; Kokta et
al., 1983).

The critical aspect ratio is. related to the fiber tensile
strength cand the fiber matrix interfacial bond strength 1: as

follows (Bigg, 1985) :

(L/D)c -25;- (1)

where, L - fiber length
D - diameter of fiber

and c - subscript for critical
Since extensive fiber damage occurs during compounding and
molding, it is desirable to increase the interfacial bond in
order to reduce the critical aspect ratio as much as possible.
Fibers between the critical aspect ratio and ten times the
critical aspect ratio provide fractional reinforcement.

The shear stresses at the interface govern the slip
process (Piggott, 1982). The shear stresses are also strongly
influenced by processing conditions and adhesion between the
fibers and the polymer matrix. Fiber breakage during
compounding and the moulding process, fiber degradation due
to processing conditions, moisture content of the fibres,
inter-fiber cling, environmental factors, orientation of the
fibers and compatibility of the fiber and the polymer matrix
are some of the factors which seem to affect interfacial shear
stresses.

A model representing an elastic matrix with uniaxially
oriented short fiber composites is given by the expression
(Piggott, 1981) :

 

¢I Ef2
0 = (<I>IEI +¢m Em)~€c - 4m .-.(2)
where c = strength

I) = volume fraction

E = modulus

9
¢=volume fraction
E=modulus
e=unit elongation
r=interfacial fiber matrix
shear stress
q -fiber aspect ratio
f, masubscripts for fiber 8
matrix, respectively
The values of r and q are major contributing factors in
determining reinforcement.
The tensile strength of short fiber-filled composites is
commonly represented as ( Bigg, 1987) :

6c: am (pm + Of ¢f 80 81 ...(3)

where, cc =tensile strength of composite
cm =tensile strength of matrix
O'f- tensile strength of fibers
(pm - volume fraction of matrix
¢f8 volume fraction of fibers
50 - efficiency factor : fiber
orientation : (80 < = 1)
81=efficiency factor : fiber
reinforcement : (0 < e < 0.95)
Random in-plane oriented fibers are most frequently
encountered in injection moulded and compression moulded

composites.

10
Theoretical predictions versus experimental data are
treated separately as sub-chapter 'F' under 'RESULTS AND
DISCUSSIONS’.

I II . COMPOUNDING TECHNIQUES

COMPOUNDING TECHNIQUES

A. W

To enhance the properties of resins, they are usually
compounded with other materials and tailored to end uses. The
process of intimately mixing the components into a
homogeneous mass is known as "compounding" (Plastics
Engineering Handbook, 1987). Compounding involves fusion of
different materials into a homogeneous mass, uniform in
composition and structure. When thermoplastics are
compounded, they are subjected to laminar shear deformations
so that the initial non-random distribution approaches some
degree of randomness. However, due to interparticulate forces
leading to particle agglomeration, adequate stress must be
exerted on such agglomerates to ensure proper compounding.
Compounding results in substantial breakage of the reinforcing
fibers. Bigg (1985), concluded that the damage done to fibers is
related to the type of compounder used.

»Techniques used for compounding are governed by the
nature of the resin and compound to be incorporated.
LWLMW

Continuous compounding provides the advantages of
economies of scale. But when the feedstock is not meterable,
batch system like the Banbury mixer would be the choice. Both
processes are widely used, however, batch processes have
certain limitations such as :

a). Potential batch to batch variation.
1 1

12

b). Greater manpower requirement.

c). Poor economies of scale.

d). Lesser output per unit time.

c). Tend to occupy more floor space.

1'). Equipment generally more bulky.

g). Set-up time high.

h). More handling equipment required.
1W:

a).B_an_b_u:y__Mj_xe_r_ It is a batch mixer consisting of

totally enclosed mixing chambers two spiral shaped rotors

providing the kneading action. Batch temperature is controlled
by steam or cooling water fed to the cored rotor. The rotors
revolve in opposite directions and at slightly different speeds.
Material is fed through a hopper and an air operated ram
confines the batch within the mixing chamber. The ridge
between the two cylindrical chamber sections assists in the
mixing; and the convergence of the rotors and the chamber
walls imparts shear.

b).C_Qn_tj_ny_Qy§__Mj_x_er The rotors and mixing action are
similar to Banbury and material is interchanged between the
two bores of the mixing section. A short feed zone leads to the
mixing zone where intensive shear occurs between the rotor
and the chamber wall. Temperature can be adjusted over a
wide range and then set constant for the production run. The
mixer is not completely self purging. The single screw and

multiple screw extruders can also be grouped under continous

mixers. A schematic view of an extruder is shown in Figure 1.

 

13

 

 

Lmumbxm ca Co 26:, £39.28 ._ 9.sz

    

 

26.5w «6.... drills .233;

_ _ _
...Pb EFII- S m b.-

"“““““‘

 

 

 

 

     

 

 

 

 

 

 

..B
I3.
2 _ 8R8.
gang
uwdm cam .33;
3a 8.:de

 

3%

 

 

14

C).T£Q_RQU_Mfll§_ They consist of two counter-rotating

parallel rollers placed close to one another with the roll axes
lying on the horizontal plane so that a relatively small space or
nip between the cylindrical surfaces exist. The rollers rotate at
different speeds to facilitate formation of sheets and the gap
between them determines the severity of mixing and the
thickness of the formed sheets. Material is deformed by friction
forces between itself and the rollers. To ensure a highly
homogeneous mix, a series of rollers are sometimes used for
continuous compounding. The rollers are heated by steam or
cooled by water to maintain the desired compounding
temperature.

d).I$md_m They work on the principle of both drag
flow and positive pumping action. The rotating screws also
oscillate between the stationary teeth in the barrel and
interrupted flights on the screw. Material movement is in an
axial direction. Due to the machines unique melting and mixing
characteristics, processing of materials can be achieved in a
much shorter length as compared to counterparts of the single
or twin screw extruder. High shear is produced in the material.
3. WM

The general melting mechanism of a polymer in a screw
of an extruder is shown in Figure 2. The melting in the screw of
an extruder starts typically in the feed section and continues
through the transition and the metering sections. The start and
end of the melting and the mechanism of melting depend

upon many factors such as type ‘of resin used, the feed

 

16

stock (pellets, granules, powder,etc.), the screw design, the
barrel profile, and the output rate. The mechanical energy
created by shear due to rotation of the screw is converted to
heat and this along with the heat provided by the external
heating system melts the material. A typical heat balance of a
corotating twin-screw extruder shows that approximately 55%
of the energy reaching the material comes from the screws and
45% from the heater bands. The shear rate "y"and the related
power "2", which is transferred to the material when the flights

are full is (Martelli, 1983) :
D n
Shear Rate, 7 =“_F9_ (Sec ‘1) (4)

where D6 = equivalent screw dia. for
the twin-screws.
n = screw speed

h = channel depth

Power transferred (Martelli, 1983),
Z: uyz V.10‘1o (Kw) (5)

where, 1.1 = coefficient of viscosity
= volume of material involved
= area of channel x
equivalent elix x

no. of flights filled

17

This is the power which is dissipated into the material as heat,
and contributes to the melting because of shear. Various
mechanisms like reverse flights, zero flights and forward
flights are incorporated in the screw design by proper
orientation of the mixing paddles to create shear at a particular
section of the screw where we want to melt the material.
C. W

Fibrous reinforcement of a polymer matrix demands
sophisticated mixing equipment which must provide extensive
intake and conveying capabilities, polymer wetting, and
dispersion of the reinforcement. The resin and the constituents
must be blended to yield a product that combines the
properties of each of the components. Some of the basic
requirements in compounding fiber reinforced thermoplastic
(FRTP) materials are (Stade, 1977) :

a). Low, controllable thermal degradation of base
polymer.

b). Optimum fiber length in the finished product.

0). Good dispersion of additional additives, eg. inorganic
fillers or flame retardant additives.

d). An economic production.
The process should provide for controlled shear, temperature
and residence time. This is to minimise material exposure to
heat, prevent degradation, and to meet product requirements.

The extrusion process is a proven economical method for
fiber reinforcement of polymers and co-rotating intermeshing

twin-screw extruders are particularly suited for these tasks

18

(Else et.al, 1985). Positive conveying, self wiping, and shear
sensitive mixing characteristics provided by the screw
mechanism satisfy requirements of reinforcement
compounding. This mechanism results in interruption of
streamline flow, which is needed to disperse both high and low
aspect ratio reinforcing agents into a suitable polymer matrix.
1W

The aim of the extrusion process is to mix the matrix and
the filler so that samples taken from any portion of the
extrudate show uniform properties. Twin-screw extruders are
superior to single-screw extruders in terms of mixing
capability and pumping. It is sometimes difficult to get stable
operation in a single-screw extruder at high throughput rates
because the extent of shear, the material temperature, and the
mixing are highly interdependent and cannot be separately
controlled (Levy, 1981; Gibbons et. al., 1987). In a twin-screw
extruder, the functions of pressure buildup and shear mixing
are basically independent and it is possible to vary each of
them by appropiate screw design and machine control. The
ability to control shear therefore makes it specifically
suitable for compounding shear sensitive materials. Further,
two stage twin-screw extruders enable a more gentle
treatment to the fibers, which are fed to the second port to
minimize machine wear and fiber breakage (Jakopin, 1982;
"Operating Manual, Baker Perkins MPC/V, " 1987).

19
2. {Iiwjn Sgrgw Extmde:

The operation of a co-rotating and counter-rotating twin-
screw extruder are shown in Figures 3 & 4. In twin-screw
extruders, the depth of the flights, their intermeshing and their
fit within the channels are of significant importance. Non-
intermeshing screws function in the same manner as a single-
screw extruder and friction is the basic factor which causes
flow of the material. When one screw penetrates into the
channel of the other, there is interaction between the screw
and the material contained in the channel. The flights of the
screw limit the rotational movement of the material round each
screw and as they continue to rotate, the materials slide within
the channels, assuming an almost linear motion towards the
end of the barrel _ thus resulting in positive pumping action.
The more the flights of one screw match with the channels of
the other screw, the more prominent is the pumping.

The drag or pressure flow does not have any significant
effect on the output or pressure at the head of the extruder.
Unlike the single-screw extruder, the maximum shear occurs at
the root of the screw and not at the barrel. The drag, pressure,
and net-velocity profiles for a twin-screw extruder are shown
in Figure 5 (Wyman,1975). The drag and pressure flows are
additive and equal to the flow of the progressing cavity to
produce a dynamic equilibrium condition. Shear on the
material is unaffected by the delivery rate (Levy, 1981) and

much less shear is produced under normal operating

2 O
CO-ROTATING

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Screw m
Section

HQWJWW

 

 

CDUNTER-RDTATI NG

//rp (r?
I: I

J VCJJ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. /

Screw Barrel l

SQCIIOII

 

 

figure 4. qunter-roteting twin-secew operation

 

 

21

00

Bo: Bo: Bo:

E023 oSmmoa . ago
Do:

 

 

 

 

05

 

 

 

 

 

 

 

 

 

 

 

 

 

§§§§§§§§§§§ 1.

22

conditions. The mixing mechanism ensures that all the
materials go through the same shear history.

In counter-rotating screws, the material is more or less
sealed in C-shaped chambers formed by the intermeshing
conjugated threads. The drag effect makes the material flow in
opposite directions. Leakage flow around the screw and leakage
within the intermeshing region due to non-conjugation (for less
shear) reduces the pumping efficiency as the factor "q" in the

equation below increases .

ie. Qmax = Qc — q, ...(6)

where, Qmax = maximum output
Qc = output without leakage
q = leakage flow
For counter-rotating screws the value of "q" is between 0.65Qc
to 0.5Qc and mixing is determined mainly by the size of the
gaps.

The situation is quite different for a co-rotating screw
where the material is not enclosed but there are passages at
the flight tips and material is transported to the second screw
in a figure '8' path. The material comming from a single
channel of one screw is divided into two streams, each of which
joins with another coming from a different channel. This
mixture is further divided at the next and subsequent
intermeshing points. The output of the screws however does
not diminish appreciably as the drag flow acts in the same

direction as the pressure flow.

23

D W

Extrusion compounding enables processing of a broad
range of products wihout hardware modification. Some of its
attractive features are uniformity of stress distribution, ease of
maintenance, self cleaning, segmental screw design, and
reproducibility of product quality even with worn parts. When
all the ingredients are meltable with similar melt
temperatures, a single stage compounding would be the choice.
However, for alloying polymers having widely different melt
temperatures or viscosity, reinforcing soft and fragile fibers or
for incorporating abrasive fillers, a two stage operation

increases the versatility of extrusion compounding.

IV. MATERIALS AND METHODS

 

 

 

 

 

i \r\\\\\\\\

 £/..... a a

 

 

 

 

 

 

26

temperature, a wide land orifice plug was selected so as to
ensure complete melting of the polymer matrix.

Paddles are the primary working component of the
agitator shaft. The motion of the paddles relative to each other
and the barrel causes the melting of the polymer, dispersion of
the agglomerates, and distribution of the ingredients.
Segmental screw design enables ease of set-up for both
forwarding (downstream) and reversing (upstream) conveying.
The conveying ability of the extruder was varied by
manipulating the relative orientation of the paddles. The
general rule followed while designing paddle orientation was

a). Conveying ability greatest for 45° paddles.

b). No conveying tendency for 90° paddles.

c). Some conveying tendency for 60° paddles.

Mixing stage 1 and mixing stage 2 are also shown in
Table 1 along with their relative conveying abilities.
Compounding temperatures were 150° C, 170° C, 190° C, and
210° C. The screw speed was 100 rpm. The processing
parameters were set through a computer connected to the
extruder.
1mm

The extrudate from the MPC/ V—3ODE was compression
molded in a Carver Laboratory Press (Model M, 25 Ton). The
internal dimensions of the mold were 6" x 6" x 1/8". Mylar slip
sheets were placed on both sides of the mold. A pressure of 3
Ton was maintained in the mold for 10 minutes at 150° C. Cold

water was circulated through the press to cool the molded

27

sheet for another 15 minutes. The sheet was finally conditioned
at room temperature for at least 48 hours before cutting into
standard sizes for tensile, impact, and flexural tests.
0.1211st

ASTM D638-86, "Standard Test Method for TENSILE
PROPERTIES OF PLASTICS," was followed. Test specimens were cut
from the molded sheets to sizes of 6" x 3/4" x 1/8" by a band
saw. These were further out to dumbbell shape type I
according to ASTM 638 in a Tensilkut Machine (Model 10-13).
An Instron Tensile Machine (Model 1114) was used to measure
tensile properties. Prior to the tests it was observed that in
order to obtain a good output in the chart recorder it was
necessary to maintain a cross-head speed of 2 in/min for HDPE
test samples and 0.5 in/min for wood fiber composites _ there
being no significant differences of tensile strength, tensile
modulus and elongations at these two strain rates. Test

conditions were as follows :

a). Load cell range : 500 lbs.
b). Cross-head speed
i). HDPE : 2 in/min.
ii).wood-fiber composite : 0.5 in/min.
c). Chart-speed : 5 in/min.
(1). Grip separation : 2 in.

Abrasive paper was used between the grips to prevent
slippage. The chart recordings were analysed for calculating

tensile modulus and elongation at break.

28

The samples consisted of different weight percentages of
untreated, heat-treated and acetylated wood-fibers with HDPE
as the polymer matrix. Maximum fiber loading was 60%
(weight percent).

ii)- ImmatJmath

ASTM D-256-84, "Standard Test Methods for IMPACT
RESISTANCE OF PLASTICS AND ELECTRICAL INSULATING MATERIALS,”
was followed. Test specimens were cut to sizes of 2 1/2" x 1/2"
x 1/8" and notched to a depth of 0.1 in. in a TMI Notching
Machine (Model 22-05). Izod impact strength was measured in
a TMI Impact Testing Machine (Model 43-02-00) using a 5
(five) pound pendulum. The machine is programmed to give
digital readout for the average impact strength & the standard
deviation. Printed results were obtained from an on-line
printer. The samples consisted of different weight percentages
of untreated, heat-treated and acetylated wood-fibers with
HDPE as the polymer matrix. Maximum fiber loading achieved
in the extruder was 60% (weight percent).

iii). W

Procedure ASTM D790-86, "Standard Test Methods for
FLEXURAL PROPERTIES OF UNREINFORCED AND REINFORCED PLASTICS
AND ELECTRICAL INSULATION MATERIALS," was followed. A 22 KPS
MTS Testing Machine was used for the flexural tests. Test
specimens were cut to sizes of 5" x 1/2" x 1/8". Flexural
properties were measured based on a three-point loading
system. The support span was 4" and support span to depth

ratio was 32 : 1. An on-line printer attached to the MT 8 Testing

29

Machine recorded the load deflection plots. The flexural yield
strength and the energy at yield were calculated at a cross-
head speed of 0.04 in/min.

The samples consisted of different weight percentages of
untreated, heat-treated and acetylated wood fibers with HDPE
as the polymer matrix. Maximum fiber loading was 60%
(weight percent).

iv). WW

Samples were cut from the molded plates representing
different fiber weight percentages. These were boiled in xylene
to dissolve the HDPE. Slides containing the wood fibers were
then prepared to project the images on a large screen with a
magnification of about 15 (fifteen). Fiber length was
determined by measuring the length of each fiber. Each datum
was a measure of the average of approximately 300 fibers.

v. W

A JEOL T-300 scanning electron microscope (SEM) was
used to study the fracture surface of tensile test specimens.
The test specimens of size 1/4" x 1/ " x 1/2", with the fracture
surface facing upwards, were glued to a circular pedestal using
aluminum paint. They were coated with gold to a thickness of
approximately 200 Angstroms in a Polaron Plasma Coater to
facilitate electron scanning. The study of the fracture surfaces

was conducted in magnification ranges from 100 to 5000.

RESULTS AND DISCUSSIONS

RESULTS AND DISCUSSIONS

A maximum fiber loading of 60°/o by weight can be
achieved in a HDPE matrix while compounding in a twin-screw
extruder. The density of the wood fiber composite increases
with fiber content and is approximately 1.20 g/cc for 60%
weight of fiber content. Each datum reported here is the mean
of at least three measurements and the error bars represent
one standard deviation from the mean.

A. EEEECLQEEIBEBIBEAIMENI

The effect of fiber treatment on the mechanical
properties of wood fiber composites was first evaluated. The
fibers studied were untreated, acetylated and heat treated.
These fibers were compounded with HDPE regrind in a Baker
Perkins twin screw extruder (Model MPC/V-3ODE). The

processing conditions were as follows :

a). Screw Configuration : SC 1(see Table 1)
b). Processing Temperature : 150 °C.
0). Screw Speed : 100 rpm.

1, -. '- ;.. . ;. ' .... , ... up... ._ .
Figures 7, 8, and 9 show the tensile strength, tensile
modulus, and elongation at break of the composite as a
function of fiber content. The tensile strength decreases
gradually with increase in fiber content. The data for
acetylated and untreated fibers showed an upper bound trend,
indicating adhesion, while heat treated fibers showed a lower

bound response, indicating low adhesion. Bataille (1987), found
30

31

 

 

 

8852:... 85% u <0
682:8 @980 n “.0
5285928 95:8 .1. om
80.24
00 :9: 29: nose E8... “.09. “.03 “.09. E09. “.0 me <08 <08 “.08 “.08 008 ...0O v
00 :9: 29: <08 <08 “.08 ".08 “.08 ...08 <08 <08 <08 “.08 “.08 ”.08 m
om 8.: awe. <08 <08 <08 <08 <08 ...08 <08 <08 <08 <08 “.08 “.08 N
om 62.93928 <08 <08 “.08 “.08 “.08 00° <08 <08 ...08 E08 “.08 .....0o P
8:38

 

 

 

 

 

32

O O I

n 2983 L8:

on ov ow o

 

 

mac: 0
. o m 8232238,. I .82

L3: 8.8: 28: o

 

 

 

 

. m 82.. 82885 n . 88

. * o . 88.
o L.

L .. 82.

 

 

a . . . ooom

(Isd) uifiuens ausuej

33

a Ema; L3:

00 8 ON

 

 

to: o
W :5: 822388. I
.3: 85858: o

.82... umuoobc: D

 

 

 

 

r cocoo—

.. ooooom

.. ooooom

 

ooooov

 

(Isd) snlnpou ausuej

34

.o .1 .. .. L. T . ace. .2. .83.:

u Ema; LB:

 

 

 

 

 

 

 

on 9. ow o
L.- - n O
c
o o
I
O o .W
m o * -o.
to: o a
.3: 832382 I
.8
.2: 83.2531. 0 00
5223855 a
.8
. IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII t
4 q . 0V

 

)IDSJQ re uonefiuola x

35

a lower bound response for tensile strength versus untreated
cellulose in a polypropylene matrix indicating no adhesion
between the two components. The data of other researchers
(Klason et.al., 1984; Woodhams et. al., 1984) show an upper
bound response, perhaps due to difference in fiber species,
treatment or processing technique. Composites of untreated
and acetylated fibers seem to have the same strength which is
higher than that of heat treated fibers. The higher tensile
strength of the untreated fibers is possibly due to retention of
most of their Iignin and natural waxes _ materials which can
aid fiber dispersion in non-polar hydrocarbon polymers
(Woodhams, 1984). The tensile modulus increases with fiber
content. It seems that there is adhesion between the wood
fibers and the polymer matrix. The fibers are able to resist the
stresses and strains set up in the polymer matrix resulting in
an increase in elastic energy and consequent stiffening of the
composite. The increase in tensile modulus is about 250% at
60% fiber weight. There is a precipitous drop in tensile
modulus at 62% for heat treated fiber _ perhaps due to
excessive fiber interaction at high fiber content. Woodhams
et.al., 1984, Beshay et.al, 1985, and Kokta et.al., 1986 did not
find such a drop at high fiber levels although it may be noted
that maximum fiber loading in their case was 40%, 50%, and
40% by weight respectively, with fiber treatments differing
from this study. Fiber treatment does not seem to have any
effect on tensile modulus so that the choice of material will

depend upon price. There is also a sharp drop in elongation

36

(Figure 9) at low filler levels. Elongation properties do not
change with fiber treatment and tend to drop to zero values
beyond 50% fiber content. It seems that there is very little
effect on elongation beyond 35% fiber content.
2. Warmth

Figure 10 represents the investigations for Izod impact
strength. A sharp drop is seen at low fiber contents, perhaps
because of voids in the sample due to volatiles generated by
oils, waxes, moisture, and by degradation of the lignin and
hemicellulose at high temperatures. The recycled HDPE matrix
has an Izod impact strength of 2.5 ft-lb/in and the average
impact strength for the fibers with different treatments
levels out to 0.7 ft-lb/in for fiber loadings beyond 20% by
weight. It seems that voids between the hydrophobic matrix
and the hydrophilic wood fibers result in propagation of
fracture even at low fiber contents.
3. HEW

The flexural yield strength of the composite seems to be
better than the polymer matrix (Figure 11). The untreated and
acetylated fibers also perform better than heat treated fibers.
The strength first decreases to a minimum and then after
reaching a maximum, again drops with increase in fiber
loading. There is some reinforcement of flexural strength for
fiber contents between 20% and 50%. This would seem to agree
with the findings of Woodhams (1984) that a critical fiber
content is required for attaining reinforcement. The maximum

value for flexural modulus is 75,000 psi. a t 60% fiber weight

 

Nu... om. ._ 09 4:303 an: m> 528% 3008.. cow. .0. 8:9...

a Ema; L8:

 

37

 

8 8 om
_ n n 0.0
0 W U 0 r m6
D ,

n. n. m o m o
. 3
mac: 0 1 m _

LB: 832382 I

- ca

.82.... 8.39; “am: 0

.85... 93095:: D

 

 

 

 

 

 

mN

o.m

(UI/CIl-ll) uifiuans 109de DOZI

38

m 2925 .3:

on on Co. ow

 

 

“no: .-
.2:lum::mxa< .-

..onE @303... “no: 0

 

 

.85”. gouache: D

 

 

.ooo.
rooom
-ooom
wooov
wooom

r 0000

 

 

ooow

(.lsd) U15U8JIS mam lanxau

39

(Figure 12). Compared to the polymer matrix, there is a 340%
increase in the modulus. Excessive fiber interaction at high
fiber contents has perhaps resulted in a precipitous drop at
62%. Fiber treatment does not seem to have any effect on the
flexural modulus. Woodhams, 1984 also found ”minor
differences" in flexural modulus for different pulp samples
including softwood pulps, hardwood pulps, or newsprint
reclaim. However, it may be noted that he added a small
percentage of stearic acid (2% by weight of pulp) to H'DPE and
used a Brabender Mixer with slightly different mixing
conditions.

4. splegtipn 91 fine: from QLQ-tzealmems

From the above results, we find that acetylated and
untreated fibers perform better than heat treated fibers for
tensile and flexural yield strength. Fiber treatment also does
not have any effect on tensile modulus, flexural modulus,
impact, and elongation properties of the wood fiber
composite.

Since untreated fibers are least expensive, they would be
most suitable for commercial applications. These fibers have
therefore been used for further evaluation.

3. W

Four screw configurations (Table 1) were examined. The
starting configuration was chosen by considering a balanced
conveying ability. The various sections of the screw and the
orientations of the paddles in the mixing section were marked

on an ”Agitator design worksheet” (Figure 13). This design

40

o o

on on

- . 1 f - 1 .
O

n 29...; been.
ov om

 

WI

 

men:
so: e3238<

.82... 339F501

.82“. 9335::

 

 

 

.. coco—

.. OOOON

u oooom

.. oooov

r oooom

r 00000

I oooon

f 00000

00000

 

 

 

(,lsd) snlnpou [OJnxalj

41

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

e s m m .Ie m giggles».
h w Ala h d a
C d u c d u 5 Mg
«awn emwz
_ _ _ _
30.0w gmhom m 30.5w r m 30.0w gmhom m 39.0w
Be Be 5 89 5 .89 .89 m $5586
.88 me .58 me m .58 me n m 8.: cm .EE me 7. x03 .953
F 33m - N mama r
.EE o . 9.3:. ..EE ON: 9.38 . .EE mmw .EE mm;
use 853 age 855
23000855 new. on? oaaoowgofi new. :97. 2389:55
_ _

 

 

anq o>_m> ..oa .023
_

.m ._ : BEE 9.88 2th

 

 

42

sheet is used for the assembly'of the twin screws of the
extruder and serves as a reference for comparing diferent
screw designs. Variations were introduced in the orientations
of the paddles in both mixing stage 1 and mixing stage 2 in
order to make changes in the conveying ability and mixing.
Screw configuration 2 (Table 1) was eliminated because of its
poor conveying ability.

The mixing stage 1 has to ensure that the polymer, which
is introduced in the first port, completely melts before it
reaches mixing stage 2. A wide land orifice plug placed
immediately after the mixing stage 1 assisted in the melting
by maintaining the filled length. The barrel of the extruder was
opened quickly after the extrusion to observe the melt in
mixing stage 1. It was seen that there was proper melting of
the polymer in all the three paddle configurations of mixing
stage 1.

The variations in the paddles of mixing stage 2 had
different effects on fiber breakage and mechanical properties
of the composite. The effects of screw configurations 1, 3 and
4 on the fiber length distribution are shown in Figure 14. The

processing conditions were as follows :

a). Screw configuration : 1, 3 and 4
a). Processing Temperature : 150 °C.
b). Screw Speed : 100 rpm.

The better mixing (low conveying ability) of configuration 1
caused maximum fiber damage and resulted in shorter fibers

for a fiber content of 50 °/e by weight. The fiber length

43

 
  
 
  
 

 

Before Compounding
— - — - Configuration 1
Configuration 3

 

...m...“ Configuration 4

 

 

Fiber Length (mm)

Figure 14. Probability density curves for fiber length_.

44

distributions for configurations 3 and 4 are almost identical.
The mean fiber length had reduced from 0.8 mm to 0.4 mm _
bringing down the- aspect ratio by 50%.

The results of tensile strength (Figure 15) and tensile
modulus. (Figure 16) show that configuration 1 is better than
configuration 3 and 4. Flexural yield strength (Figure 17) and
flexural modulus (Figure 18) show the same trend. We may
again note that in the case of flexural yield strength, the
trend is first a decrease to a minimum, then increase to a
maximum and finally decrease with increasing fiber content.

The data under this processing condition seem to
indicate that shorter fiber lengths impart higher strength to
the composite, which is generally contrary to expectations. We
may however note that Configuration 1 provides better
dispersion and tends to enhance fiber matrix interaction with
resultant shorter fiber length. Recalling equation 2 (Piggott,
1981), it was expected that the strength of uniaxially oriented
short fiber composites would increase with fiber length and
fiber-matrix adhession. However, from the above analysis
fiber adhesion seems to be a more dominating factor than fiber
length for imparting strength to the composite. McNally et.al.,
1978, found that better dispersion results in a stronger
interface, which might lead to a stronger composite even
though the fiber length is degraded. This may be due to better
wetting of the fibers and the absence of microvoids. Bitaille,
1987, also found an increase of elastic modulus with mixing

time using a Brabender Mixer to mix 20 % cellulose with

45

ov

a 2925 :5:

ON

 

 

 

 

ccozoczmacou I
McozPsmccou o

_ 5395928 U

 

 

wooe—

tooow

tooom

 

ooov

 

. oeem

(lSd) Lnfiuans ausual

46

 

G. on.— ..322 33333 £2238 2.33“ me um um Guuuu an 3&3

on

ov

a 2935 .2:

ON

 

 

 

 

v 5395928 I

n 5395923 0

_ 5:953:89 n.

 

 

 

l

 

 

noooom

vooooo_

noooom_

rooooow

roooomm

rooooom

oooomm

(,lsd) snlnpou ensue;

47

O

a 25.53 been.
ov

om

 

 

 

 

v5.2.5928 I
m5305328 o

_ 5.2.5328 D

 

 

loco—

nooom

nooom

nooov

 

ooom

 

. oooo

(.lsd) utfiuans Dial/l leJnxau

48

O

on

.‘

m3 .0: m 3 .

ov

a see; 23.:

ON

 

 

 

 

v 5395928 I

n 5395928 0

_ 5395928 U

 

 

n

noooow

noooov

nooooo

fooooo

 

 

(lsd) snlnpouJ (anxaL-l

49

polypropylene at 190 °C. It seems that this complex behaviour
needs more indepth study so as to characterise the same.
C W

We may express the fiber length in terms of moments of
distribution similar to the manner a polymer chemist

describes molecular weight distribution.

ZNiLi
Ln um (7)
ZNiLiz
and Lw 'W (8)

The result is shown in Table-2, where Lw/Ln is the
polydispersity. The average width of the fibers was 0.03 mm.
D. W

Compounding temperature is an important design factor
for processing. Thermal degradation of the fiber sets limits to
the processing temperature. Lignin starts to degrade at 200 °C.
and most natural fibers begin to lose their strength at 160 °C.
(Cruz-Ramos, 1986). The effect of compounding temperature
was therefore studied in the range of150 °C. to 210 °C..

1. WM

Processing conditions were as follows :

a). Screw Configuration :SC1

b). Screw Speed : 100 rpm.

0). Compounding temperature: 150 °C., 170 °C., 190 °C.,
and 210 °C.

d). Fiber Type : Untreated Aspen

50

 

 

 

mm. P 5.0 and .5 22859200

em. F 86 and m 22:83:50

em; 5.0 med F 83939200

mm; mm; mm; 9.658200 928
5:3 :55 2.. 55 5 ad:

 

 

 

 

 

 

51

Tensile strength and tensile modulus are both highest for

a processing temperature of 150 °C. (Figure 19 and Figure 20).

aElezuLaLstreuotLaanJamLmodulus

Processing conditions were as follows :

a). Screw Configuration :SCS

b). Screw Speed : 100 rpm.

0). Compounding temperature: 150 °C., 170 °C., 190 °C.,
and 210 °C.

c). Fiber Type : Untreated Aspen

Flexural strength and flexural modulus seems to be highest for
a processing temperature of 170 °C. (Figure 21 and Figure 22).

The results seem to indicate that tensile and flexural
properties are sensitive to compounding temperature. It is
further observed that for better tensile properties, a longer-
mixing time (screw configuration 1) requires a lower
compounding temperature. A shorter mixing time (screw
configuration 3) requires a higher compounding temperature
for better flexural properties. The composite also tends to
darken as function of time and there seems to be
thermodegradation. Two competing mechanisms may be present
_ cellulose degradation and increase of adhesion.
E W

The fracture surfaces for tensile specimens compounded
with screw configuration 1 were examined. Figure 23 shows a
clean surface at 3500 magnification (untreated fibers, 150 °C.,
32% fiber). Scanning electron micrographs at 1500

magnification show adhesion for untreated fiber at 170 °C.

 

52

 

 

 

 

... .. J. -. -._. mm. .0 -. .=. .u...-=..qo O o .
n 2925 28:
on om 9» on om o. o
- n b b n u o
“Yo—N I 0.0: O 1000—
u.o2 I 98. a
O ..ooow
..ooom
o
D ..ooov
a q u u - 1 GOOD

 

(.lsd) Lnfiuans ausuei

53

3 ..= .

a 592; :5:

...q.

o 03..

 

 

 

 

 

 

 

on 00 cm 0v on ow o_ o
L n b h n n o
1 W .. 00000—
1 % u OOOOON
L u. om. I 0. cm. D .. 00000»
d d d d u -

 

(lsd) snlnpou eusuel

54

a EB; 22.:

 

 

 

 

 

 

 

 

8 em 8 on 8 o. o
n n h n n o
1 u.o_N C 90$— . loco—
H
u. o: O u. om. D a
.. r 88 m
J
m.
.. - coon M
m.
p
t . 89. m
J
a
U
1 m . ooem .m.
u.
\d;
. . eeee new
econ

55

 

 

 

 

 

. em - ..= a e. _. .2. .u... ...u. . 3..
u see; .8:

8 cm ov on on o. e

- n u u n o
. 8.5 o 8.8. I % -88.

92. o 98. a
1 IOOOON
. * .88n
- reooov
- w W .888

eeeoe

 

 

 

(lsd) snlnpou [cunxelj

56

(Figure 24) and lack of adhesion at 190 °C.(Figure 25) and
210°C.(Figure 26). The fiber content was approximately 30% by
weight. Heat treated fibers at 1500 magnification (Figure 27)
also showed lack of adhesion at 150 °C.. The fiber
concentration was 35.8% by weight. Acetylated fibers at 28%
by weight show some adhesion at 2000 magnification (Figure
28). Other micrographs at magnifications of 350 and 200
indicate a high degree of fiber pull-outs with consequent low
degree of adhesion for fibers processed at 150 °C.

A comparison of the micrographs for 30% acetylated,
30% untreated, and 33% heat treated fibers, which were
processed with screw configuration 1 (Table 1) do reveal that
there is considerable adhesion of HDPE on acetylated fibers
and very little adhesion on heat treated fibers. The fracture
surface of the composite with untreated fibers however
indicate a large degree of fiber pull outs. Their corresponding
tensile strengths of 3900 psi., 3700 psi., and 2150 psi. would
perhaps correlate with the micrographs. In order to make a
conclusive statement regarding this complex phenomenon of
adhesion, a more indepth study of the fracture surface using
SEM would be necessary. However, the large degree of fiber
pull outs and their clean surfaces seem to indicate the
presence of voids as one the contributing factors in crack
propagation. The above studies also seem to indicate the
potentiality for increasing the strength of the wood fiber

composite by better bonding between the hydrophhobic HDPE

i—
r"

' 9‘

\!

 

15KU X3’506

Fi 1' 2 E " n fi er QC, 3;: 4

 

Figure 24. SEM ;; 39 % untreated fibers, 150 O—Q, fig 1

15KU XI’SQB

 

10Mm 02004'

Fioure 26 EM "2 % nre fiber 15 92.59.;

~
109m 000013

 

Figure 22, SEM ;; 38,5 % heat tregteg fiber; 150 m

 

Figure 28. SEM :; 28 20 eeetylated fibers, 15(2 QC, SC 1

 

60

and the hydrophilic aspen fibers and reduction of voids.
F. W

There are several theories in the literature that can
predict tensile strength and tensile modulus of filled polymers
but there seems to be none that can predict impact strength of
filled polymers.
1. W11

Landon (1977) and Bigg (1979, 1987) showed upper bound
and lower bound relationships between the tensile strength
and the volume fraction of the filler. The upper bound curve
indicates adhesion while the lower bound curve indicates little
or no adhesion. The lower bound response can be expressed by

the mathematical model (Nicholais et.al.,1973) :
- 1 - b 9
cc op( a (b ) ...( )

where, 0c - tensile strength of composite
cp a tensile strength of polymer
a =- constant related to stress conc.
b - constant related to geometry of
filler

¢ =- volume fraction of filler

c
or, 39- 1-arpb ...(10)

61
or, In [1 - 03-]: In a+b In it) ...(11)
“P

Bigg (1987) has reported that for spherical particles
having no adhesion to the polymer matrix, 'a' has been found to
be equal to 1.21 and when there is some adhesion 'a' is smaller
than 1.21. The value of 'b' is 2/3 for random fracture of a
composite and 1 when failure is by planar fracture. The
experimental data for heat treated fibers fitted quite well
(Figure 29) to equation (11) although the values of 'a' (1.68)
and 'b' (1.032) obtained from the plot seemed to deviate
slightly from the limits reported by Bigg. The coefficient of
correlation is 0.996 (Table 3). Thus the model for heat treated
fibers (Table 4) is :

oc=0p(1-1.68¢1'032 ) ...(12)

Data for untreated and acetylated fibers deviate from this
model (Figures 30,31), indicating some adhesion between the
fibers and the polymer matrix.

Thus it can be concluded that the model at equation (9),
which assumes no adhesion between the filler and the polymer
matrix, fits well to the tensile strength curve for the heat
treated fibers. Our assumption that there is lack of adhesion
for heat treated fibers agrees well with this model. For
acetylated and untreated fibers, the model indicates better

adhesion and therefore higher tensile properties.

62

35:5
8 . o - o. T m ... -

v.7 mé. mé-

db

.-

v

F
l

 

 

I-
III

 

 

 

I'

i

Hea trea e

DP m rix

in

63

 

 

,o' 0‘ o o. .0 o g‘ ‘0 0; 01'
Will
StatWorksP‘ Data ANOVA Table
Data File: TheoryData(T)
Sum of Deg. of Mean

Source Squares Freedom Squares F-Ratio Prob>F
Model 0.744 1 0.744 240.116 0.003
Error 0.006 2 0.003
Total 0.750 3

Coefficient of Determination 0.992

Coefficient of Correlation 0. 996

Standard Error of Estimate 0.056

Durbin-Watson Statistic

2.699

64

StatWorksT“ Data Coefficients
Data File: TheoryData(T)

 

Variable Std. Err. t

Name Coefficient Estimate Statistic Prob > t
Constant 0.517 0.079 6.526 0.019
lnfi(HT) 1.031 0.067 15.496 0.003

Fi

65

 

 

 

 

re
n rea ed

1
11 fi

_ CW .h'- O\"O—°-. on

f eor wih nil
r ' HDPE rix

1‘

n

h

-3

-1 -‘| -1

-2.0

lnfi(UT)

66

.p.

65:.

m._.-
n

I
I-

I
.-

I
I.

 

 

 

 

 

nil

with
DP m trix

1'

h

67
2.I.ensile_m9.dulus
The frequently used models to predict modulus are
applicable only at low filler concentrations (Einstein, 1906;
Mooney, 1951, Frankle - Acrivos, 1967). Nielson, 1970,
introduced the concept of geometry of the filler particle while

Guth's model (1944) is an extension of Einstein's model ie.

n*-nl1+2.5c] (13).

to take into account interparticle attractions at higher filler
concentrations. In Einstein's model '11“ and 'n' are the viscosity
of the emulsion and solvent and 'c' is the volume fraction.
Replacing the viscosity 'n' by the tensile modulus 'E', we obtain
Einstein's model for computing modulus for materials filled
with low concentrations of non-interactive spheres. Although
the above theories give some insight into the behaviour of the
elastic modulus of polymer filled systems, a satisfactory
treatment of the strength behaviour of composites reinforced
with rigid fillers has not yet been developed (Leidner et. al.,
1974)

Besides the independent action of the filler particles,
Guth considered the mutual interaction of pairs of spheres and

generalised Einstein's viscosity law as follows :

11* - n [1 + 2.5c +14.1c2] (14)
or, E' - E [1 + 2.5c +14.1c2] (15)

68

It is proposed to modify Guth's model for wood fibers as

follows :

Ec . Ep [1 ,t at) + 612] ...(16)
or, Er-1+a¢+b¢2 ...(17)

E .
where, Er - 2% - relative modulus

The coefficient 'a' is a geometric correction factor and is
equal to 2.5 for non-interactive spheres (Einstein, 1906). The
coefficient 'b' is a correction for higher filler concentration.
Using the experimental values in the above equation, we
get the curve fits for relative modulus versus fiber content for
heat treated, acetylated, and untreated fibers as shown in
figures 32, 33 and 34. The corresponding coefficient of
correlation are 0.962, 0.994 and 1.000, respectively. This
shows a very close fit for the second degree polynomials. A
”Statworks" program was used to evaluate the above. The
values of 'a' and 'b' obtained from Tables 5, 6, and 7, give the

following mathematical model :

1. Heat-treated : Er - 1 + 1.121 4) + 7.694 112 ...(18)
2. Acetylated : Er - 1 + 3.201 e + 6.253 e2 419)
3. Untreated : Er - 1 + 2.712 4 + 2.775 e2 ...(20)

More data may be required to perfect the above models in view

of the high standard error.

69

com

0
IPO
V

com

d-

oze. x :15
8“
h

dl-O

00—-

 

 

Fl

 

 

UJ ’- t Y—VIF'A

mod 1

h or wi h ensil
HDPE matrix

f
r

in
f'

11

1'6

1
Hea

70

 

 

 

Fi
Ac

1'

la

IT

 

UJh |1—\'<0A

i
11 fi er

f r wi h t nsil
' HDPE m rix

 

-1

400

300

200

100

-100

fi(AC) x 101-3

71

com

9...: x :3:
8w
-

co..-

 

 

com
m

 

 

[JJ 5. .V-VDF-A

4

Pi

72

 

T l f n
(Tegsile medium)
StatWorkslm Data Coefficients
Data File: Modulus
Variable Std. Err. t
Name Coefficient Estimate Statistic Prob > t
Constant 0.077 0.456 0.168 0.875
fi(HT) 1.121 4.447 0.252 0.816
fi(HT) "2 7.694 9.118 0.844 0.511

73

 

T l 1 r
11 il l
StatWorksm Data Coefficients

Data File: Modulus
Variable Std. Err. t

Name Coefficient Estimate Statistic Prob > t
Constant 0.016 0.149 0.108 0.921
fi(AC) 3.201 1.800 1.778 0.218
fi(AC) "2 6.253 4.800 1.303 0.323

74

1 r n
(Tensile meeeles)

StatWorksW Data Coefficients
Data File: Modulus

 

Variable Std. Err. - t

Name Coefficient Estimate Statistic Prob > t
Constant 0.003 0.027 0.092 0.930
fi(UT) 2.712 0.266 10.207 0.002
fi(UT) "2 2.775 0.528 5.254 0.012

VI. SUMMARY

SUMMARY

A. W

Composites with recycled HDPE as the polymer matrix
and acetylated and untreated aspen fibers as fillers are
superior to heat treated fibers in terms of tensile and flexural
yield strength. Untreated fibers may therefore be the choice for
a composite with recycled HDPE because of their lower cost.
Fiber treatment does not seem to have any effect on tensile
modulus and elongation properties. Data for impact strength
may be indicative of voids between the wood fibers and the
HDPE matrix. A critical fiber content of 20% by weight is
required before reinforcement occurs for flexural yield
strength. The mechanical properties of the composite are
sensitive to screw configuration and temperature. Screw
configuration 1 (Table-l) with a processing temperature of 150
°C. imparts the best overall strength to the composite although
it has a longer mixing time and produces maximum damage to
the fibers. Dispersion of fibers seem to be a dominating factor
in these experiments. Scanning electron micrographs seem to
indicate that recycled HDPE has better adhesion to acetylated
and untreated fibers than to heat-treated aspen fibers at 30%
fiber contents. However, more research is required in this area
to make any conclusive statement. Theoretical predictions for
tensile strength seem to agree with experimental data.

Mathematical model for predicting tensile strength and tensile

75

77

5. The composite has to be finally evaluated in terms of

commercial applicability.

VII. APPENDD(

APPENDDC

A W

1.Composite material : Two or more materials combined for
suitable structural applications are known as "composite
material".

2.Co-extrusion : The extrusion process where two melt
streams or a polymer melt and fiber are combined in the die to
make an extrusion of two materials.

3.Co-rotating screws : Two screws rotating in the same
direction . either both clockwise or both counter-clockwise.

4.Counter-rotating screws : Two screws rotating in the
opposite direction _ either both clockwise or both counter-
clockwise.

5. Die : An orifice used to shape a plastic melt stream in the
extrusion process.

6. Drag flow : Mechanism by which material is caused to flow
in single screw extruder and build pressure. The polymer must
wet both the screw and barrel to produce drag flow effects.

7. Extruder : A machine which accepts solid particles or
liquid feed and conveys it through a surrounding barrel by
means of a rotating screw and pumps it under pressure
through an orifice.

8. Extruder Length to Diameter Ratio : L/D = The distance
from the forward edge of the feed opening to the forward end
of the barrel bore divided by the bore diameter and expressed

as a ratio where the diameter is reduced to 1 (one).
78

79

9. Extruder size : The nominal inside diameter of extruder
barrel.
10. Extrusion : Process of making a product by forcing material
through a die orifice.
ll. Helix angle : It is the angle the flight of the screw makes
with a plane perpendicular to the screws axis while it winds
around the roots of the screw. The elix angle determines the
pitch of the screw.
12. Flight Pitch : Distance in an axial direction from the center
of a flight at its periphery to the center of the next flight.
13. Shear rate : The rate at which a material is undergoing
deformation in response to a shear stress.
14. Shear Stress : The force per unit area applied to a planar
element inducing shear movement of a material.
15. Single -screw extruder : Extruder using a single conveying
screw to pump plastics.
16.Twin- screw extruder : Extruder using two intermeshing

screws to convey plastics.

VIII. BIBLIOGRAPHY

BIBLIOGRAPHY

Bataille, P., L. Ricard, and S. Sapieha. "Properties of Cellulose
Containing Polypropylene, ANIEQ. 791-793(1987).

Belmaris, H., A. Barrera, E. Castillo, E. Verheugen, and M.
Monjaras. "New Composite Materials from Natural Hard

Fibers," W20. SSS-5610981)-

Beshay, A, B.V. Kokta, and C. Daneault. "Use of Wood Fibers in
Thermoplastic Composites II : Polyethelene," Eelym,

magma, 361-271(1985).

Bigg, D.M. "Mechanical, Thermal and Electrical Properties of

Metal Fiber-Filled Polymer Composites," W L9,
1188-1192 (1979).

Bigg. D.M. "Effect of Compounding on the Properties of Short
Fiber Reinforced Injection Moldable Thermoplastic

Composites," W.i 20-28(1985)-
Bigg D.M. "Mechanical Properties of Particulate Filled
Polymers." W8. 115-122(1987).

Cloud, P. and R.E. Schultz. Plegties Werle, 33(1975).
Coran, A.Y. and R. Patel. US Pat. 4,323,625(1982).

Collyer, A.A. and D.W. Clegg. "An Introduction to Fiber
Reinforced Thermoplastics," in h ' ri

MW. D.W. Clegg and Collyer, eds.,

Elsevier, New York, 2-25 (1986).

Cruz-Ramos, C.A. "Natural Fiber Reinforced Thermoplastics," in
fRinfre Thrm ...

Clegg and A.A. Collyer, eds., Elsevier, New York, 65-81(l986).

Dahl, W.S. Weee Elem, Mercury Press, Northampton,
England(1948).

Einstein, A. Ann, der Ehy,1_2, 289(1906).

80

81

Eise, K., J. Curry, and J. Farrell. "Processing of Reinforced

Plastics Using Multi-Screw Compounders," W
351,21, 497-501 (1985).

Frankel N.A. and A. Acrivos. W21, 847(1967).
Gautts, R.S.P. and MD. Campbell. "Cempegije", 19, 228(1979).

Gibbons, J. A., D. Grymberg, and H. Helmy. "Extrusion," in
W 219- 220 (1986- 87)

Goettler, L.A. US Pat. 4,376,144(1983).

Guth, E. My” 4, 4626(1970).

Jakopin, St. "Advances in Compounding Techniques for GRP,"
W09”)-

Kleson, C., J. Kubat, and H..E Stromvall. "The Efficiency of
Cellulosic Fillers in Common Thermoplastics Part I. Filling

without Processing Aids or Coupling Agents," Weds;
mm, 159- 18(1984).

Kokta, B.V., R. Chen, C. Daneault, and J. L. Valade. "Use of Wood

Fibers in Thermoplastic Composites," W4,
229-232 (1983).

Kokta, B.V., C. Daneault, and S. T. Sean. "Use of Grafted
Hardwood Pulps in Thermoplastic Composites, Eelym, Blast,

W252, 1274740986)-

Landel, R.E. "Problems encountered with Conventional Fiber-

Reinforced Composites," in Adhesieg in Qellnlesie and Weed-
W J.F. Oliver ed., Plenum Press, New York,

149-166(1981).

Leidner, J. and RT. Woodhams. "The Strength of Polymeric

Composites Containing Spherical Fillers," f li
£91,, _1_8_, 1639-1654(1974).
Levy, s. 213511;; egtggsien Ieeheelegy Handbook. New York, 1-

19 (1981).

82

Lightsey, G. R. "Organic Fillers for Thermoplastics," in 119110121;

WWW CHE Carrahar
Jr. and L..H Sperling eds, Plenum Press, New York, 193-

211(1981).

Majali, A. B., M..R Shah, and ND Patil. "Characterisation of Wood
Polymer Composites (WPC) as an Engineering Material,"

'11 011 Lat-.1 .,0“01 001,331-

335(1979).

Marra, G. G. "The Role of Adhesion and Adhesives in the Wood

Products Industry." in MW
AW. R. H. Gillespie ed., Noyes Publications,

New York, 2-9(1984).

Martelli. F.G. W
Van Nonstrand Reinhold Company, New York, 66-71(1983).

McNally, D, W.T. Freed, J.R. Shaner, and J.W. Sell. "A Method to

Evaluate the Effects of Compounding Technology on the Stress
Transfer Interface in Short Fiber Reinforced Thermoplastics,"

Eelym, Eng, Sei“ 18, 396-403(1978).
MEIR—21351125.. "Materials 1986," 43-7l(1986).

Medan—Elm "Resins, 1988 : A Year of Decission," 63-
105(1988).

Mooney, M4 M0. 162(1951).

Nicholais, L. and L. Nichodemo. "Strength of Particulate

Composite." WM“ 1.1. 4690973)
Nielson, LE. ,1, Appl, Phy,, A, 4626(1970).
"Qeetatt'ng Manna], Baker Perkins, mega," (1987).

Otey, F.H. "Current and Potential Uses of Starch Products In

Plastics." Eelym, Blast, Teehnel, Eng,,1, 221-234(1976).
Wm "Recycling Plastics," 2(1988).

83

Pal, S. K., S. K. Sanyal, and RN Mukherjee. "Natural Fiber
Reinforced Polymer Composites - a Short Review, " Ma;

mammal 30- 33(1984)

Piggctt. M.R. MW Pergamon.
Toronto, 168(1981).

Piggott, M.R. "Fiber Matrix Interactions," W3,
179-188(1982). ~

Plastics Engineering Handbook, "W
Handling. 8380937)-

Powell, P.C. W Chapman and Hall, New
York, (1983).

Ramirez, A.P., and AS. Solis. "Development of a New Composite
Material from Waste Polymers, Natural Fiber, and Material

Fillers," mm 29, 2405-2412(1984).

Rowlands , R..,E RP. Van Deweghe, and G.P..Krueger "Fiber-

Reinforced Wood Composites, " W,1_8_, 39-
57(1986).

Selke, SE, and C. Lai. "Plastics Waste and the Environment," in

Beeyeled Blames ; Develepmeat aag Applieatiens, 7-15,
Christopher C. Lai, Kit L. Yarn and Susan E. Selke, eds., School
of Packaging, Michigan State University(l987).

Semsarzadeh, M.A. "Low Cost, High Strength, Hard Cellusosic
Fiber Reinforced Polyester Structures and Units," W

W 2A. 323-334(1985).

Semsarzadeh, M.A. "Fiber Matrix Interactions in Jute Reinforced

Polyester Resin," Pelymer Cemeesite§,1, 23-25(1986).

Sheldon, R.P. Cemeesite Pelymerie Materialg, Elsevier, New York
(1982).

Shephard, P.D., F.J. Golemba, and F.W. Maine. "Mica as a
Reinforcement for Plastics," in 'l R inf r m
a]_aeti_es_," R.D. Deanin and NR. Schott, eds., American Chemical
Society, Washington, 41-51(1974).

84

Soule, EL. and HR. Hendrickson. "Bark Fiber as a Reinforcing
Agent for Plastics," W 1Q, 17-22(1966).

Stade, K. "Techniques for Compounding Glass Fiber-Reinforced
Thermoplastics," W.,11, 50-57(1977).

Theberger, J. "Recent Product Advances in Thermoplastic

Composites," Wald. 41-53(1981).

Wall, D. "The Processing of Fiber Reinforced Thermoplastics
Using Co-Rotating Twin Screw Extruders," ANIIIEQ, 778-
780(1987).

Westhoff R.P., F.H. Otey, C.L. Mehltretter, and C. Russell. "Starch-
Filled Polyvinyl Chloride Plastic-Preparation and Evaluation,"

Ind, Eng, Chcm, firm, Bcs, Dcv.,1_1, 123-125(1974).

Willis, S. and T..S King, "Surface- Modified Organic Particulates

as Extenders in Thermoplastics," Plastics Compounding, 4 46-
50 (1981).

Woodhams, R.T., G. Thomas, and D.K. Rodgers. "Wood Fibers as

Reinforcing Fillers for Polyolefins," Eclym. Eng. Sci,, 2_4_, 1166-
l 171 (1 984).

Worthy, W. "Wide Variety of Applications Spark Polymer
Composites Growth," SLéLEI‘L Chicago, 7-l3(1987).

Wyman C.E. "Theoretical Model for Intermeshing Twin Screw
Extruders: Axial Velocity Profile for Shallow Channels," Eclym.

W41. 600-611(1975)-

Yam, K.L. "Uses of Recycled Plastics in Composite Materials," in

Rccyclcd Plastics ; chelcpmcnt and Applicaticns, C.C. Lai,
Yarn and SE. Selke, School of Packaging, eds., Michigan State

University, 81-90(1987).

Youngquist, J.A. "Research Progress in Wood Based Composite

Products," AIQHE Sympcsicm Serics, 12, 79-87(1983).

.54

"Illlllllllllllllll