 

U
a! ‘lﬁf‘ﬁ: ’
' _:..z_1".u:t_
.. c {new 55.
u: 314:;

 

. '53 . =2
'ﬁy'rﬁ 3 H; ~ ' ~"‘ ' ‘ Y‘ . ‘ . " ‘ i: " r" V . %%§EH
:rf—‘viﬁs 1. - up» 3 . 3. 1" '\ 37 r} .3, _ ‘1 1' , . . ‘
girl-$553,. N, L , 3 ' ' "an“
-~.~»::%tv&: u‘ ‘
v m‘i

‘w.

4.
ﬁgy'ﬂrﬁ‘l
. ”.3. L‘ .-

1.

‘3%

 

.

w
u
r

“”3. :11. .3
mm. .....~

' It
.3.“ .
3 ~ 54.. ”$5
"2. ‘2“ .. “.913 2&5...“ #4.. .5 ~ :9: a"
' - .4 4.-.. .u‘ -.‘ H “‘M‘ ‘ ~-
~ ‘ wt w.~.- :" -F¥N v. " M
v ' I: . v ‘ . u A
“A. I \ I
0:137”: ,3: ; » '

THESE:

This is to certify that the

thesis entitled
Fracture and welding of in{ection molded
polypropylene and short 9 ass fiber reun-

forced polypropylene

presented by

Feng Cao

has been accepted towards fulﬁllment
of the requirements for

Master degree in Science

4

l/ ./ /

a
l t

f
Dr. Andre Lee

 

Major professor

Date 14-5-91!

 

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

 

 

 

lllllllllllllllll

 

 

LIBRARY
Mlchlgan State
University

 

 

 

PLACE IN RETURN BOXt oremove ethle checkoufrom your record.
TOA ID FINES return monor orbdore date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU le An Afflrmetlve Action/Equel Opportunity Inetitulon
WMt

FRACTURE AND WELDING OF INJECTION MOLDED POLYPROPYLENE
AND SHORT GLASS FIBER REINFORCED POLYPROPYLENE

By

Feng Cao

A THESIS
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Materials Science and Mechanics

1994

ABSTRACT

FRACI'URE AND WELDING OF INJECTION MOLDED POLYPROPYLENE
AND SHORT GLASS FIBRE REINFORCED POLYPROPYLENE

By

Feng Cao

The fracture toughness, Kc, was found to increase with ﬁbre content for notched
compact tention specimens in short glass ﬁbre reinforced polypropylene (GP-PP) at low
temperature. When ﬁbre content is low, (from 5% to 10% w/o, weight percent), Kc
increase quickly and the toughening mechanism is proposed to be that of crack pinning.
However, the fracture toughness increases at a slower rate when ﬁbre content is above
10% wlo. In this case, the toughening mechanism is best explained by the model of crack
deﬂection. Furthermore, when the ﬁbers are arranged transverse to the notch, we
observed an higher higher Kc value. In the welding of PP to GF—PP, signiﬁcant interfacial
strength is developed at temperature higer than 167 0C. The welding time aﬁ'ect not only
the strength but also the failure behavior of the joints. The failure occurred all through the
interface to a small part in the center of the interface as welding time increased. It was
found out that the strength of PP to GF-PP and GF-PP to GF-PP butt-joints could reach
ultimately reach that of the bulk PP.

To my parents and to Jie

ACKNOWLEDGMENTS

Thanks to my advisor, Prof. Andre Lee for his support, patience and guidance.
Thanks to Himont company for providing the experimental material, and to Prof. James
Lucas for using the Instron machine in his laboratory and for his many helpful

suggestions.

IV

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES

CHAPTER 1
INTRODCUTION

CHAPTER 2

Toughening Mechanisms of Short Glass Fibre Reinforced
Polypropylene

2.1 Introduction

2.2 Background

2.2.1 Toughening mechanisms in particle ﬁlled resin-based composites
2.2.2 Linear Elastic Fracture Mechanics

2.3 Experimental Procedure

2.3.1 Determination of ﬁbre length and arrangement in the thickness
2.3.2 Fracture Toughness Testing

2.4 Experimental Results and Discussions

2.4.1 Fibre Length Distribution in GF-PP

2.42 Fibre arrangement across the thickness of molding plaques
2.4.3 Mechanisms of toughening in short GF-PP

2.4.3.1 Mechanism of Energy-absorption in PP

2.4.3.2 Low ﬁbre content of GF-PP

2.4.3.3 High Fibre Content of GF-PP

2.4.4 The effect of ﬁbre arrangement on the fracture toughness

vii

viii

l2
14

14
l7
17
24
26
31
35

2.5 Summary
CHAPTER 3

Welding of PP and Short Glass Fibre Reinforced PP
3.1 Introduction

3.2 Background

3.2.1
3.2.2
3.2.3
3.2.4
3.3

3.3.1
3.3.2
3.3.3
3.4

3.4.1
3.4.2
3.4.3
3.5

Strengthening Mechanisms of Short Fibre Reinforced Thermoplastics
Fracture mechanisms of Semi-crystalline polymer

Fracture Mechanisms of Short Fibre Reinforced Thermoplastics
Welding of Thermoplastics and Composites

Experimental Procedures

Material

Welding Processing

Tensile tests

Experimental Results and Discussions

Tensile Properties of Bulk PP and GF-PP

Welding Parameters on Strength Development

Tensile Properties of Butt-Joints of GF—PP/GF-PP

Summary

CHAPTER 4

CONCLUSIONS

BIBLIOGRAPHY

93

39

41
42
45
47

50
5 l
54

54
56
82
89

91

List of Tables

Table 1: Fracture toughness, Kc, of notched CI‘ specimens of PP and GF-PP.
Table 2: The Results of ASTM tensile specimens of PP and GF-PP.

Table 3: The peak stress of welded joints of PP/30%GFRe-PP with various welding time
and temperature.

Table 4: Results of tensile test of GF—PP/GF-PP butt-joints.

List of Figures
Figure 1: Schematic representation of the process of crack pinning.
Figure 2: Schematics of typical crack deﬂection: (a) tilt, and (b) twist of the crack front
Figure 3: A center-cracked panel of inﬁnite plate loaded at the remote boundaries.

Figure 4: Technique for cutting samples to detect ﬁbre orientation in injection molded
short GF-PP plaque

Figure 5: (a) shows how to cut CT specimens from the received injection molded PP and
GF-PP plaque and (b) the geometry and dimensions of CT-specimen according
to ASTM.

Figure 6: Illustrates the distribution of ﬁbre length (mm) in the injection molded short
GF-PP by number percent: (a) for 5% (wlo); (b) for 10% (wlo); (c) for 20%
(wlo) and (d) for 30% (wlo) GF—PP.

Figure 7: Fibre arrangement in YZ plane, (a) of 5% (wlo) GF-PP, (b) of 10% (wlo) GF-PP,
(c) of 20% (wlo) and (d) of 30% (wlo) GF-PP (X50).

Figure 8: Injection molded induced layer structure in short GF-PP plaque. S indicates side
region and C-center region, schematically.

Figure 9: The deformation of notched CI‘ specimens of PP and GF-PP under static load
by load-opening displacement, 1 for PP, 2 for 30% (wlo) GF-PP with T-notch.

Figure 10: Fracture toughness, Kc, of notched CI' specimens of PP and short GF- PP
under static load and at low temperature (-87 °C).

Figure 11: SEM micrograph of fracture surface of notched CT specimen of PP near the
notch root showing dimple pattern of the deformed matrix.

Figure 12: SEM micrograph of fracture surface of notched CT specimen of 5% (wlo) GF-
PP with T-notch near the notch root.

Figure 13: SEM micrograph of fracture surface of notched CT specimen of 10% (wlo)
GF-PP with T-notch near the notch root.

Figure 14: SEM micrograph of fracture surface of notched CT specimen of 20% (wlo)
GF-PP with T-notch near the notch root.

Figure 15: SEM micrograph of fracture surface of notched CT specimen of 30% (wlo)
GF-PP with T-notch near the notch root.

Figure 16: Relative toughness predictions from crack deﬂection model for rod-shaped
particles of three aspect ratios.

Figure 17: Schematics to show ﬁber arrangement in T-notch (a) and L-notch (b).

Figure 18: Characteristic fracture surface of 20% (wlo) GF-PP with L-notch
shows large number of ﬁbres were arranged parallel to the notch.

Figure 19: The eﬂ'ect of a ﬁber on the matrix deformation.

Figure 20: The tensile stress and shear stress distribution along a ﬁbre.

Figure 21: Schematic steps of crazes formation in a semi-cystalline polymer structure
Figure 22: Model of fracture mechanism of short ﬁbre reinforced therrnoplastics.
Figure 23: Molecular model of healing in polymers.

Figure 25: Schematic of the welding tool and equipment.

Figure 26: Thermal history of welding process.

Figure 27: Load-displacement curves for injection molded ASTM tensile bars of PP and
short GF-PP with various ﬁbre content

Figure 28: SEM micrograph of characteristic failure surface of short GF-PP 30% (wlo) of
ASTM tensile test, viewed down on the surface.

Figure 29: Plot of peak stress of welded PP [30% (wlo) GF-PP bars with various welding
times and temperatures by tensile test.

Figure 30: Load-displacement curves for PP/30% GF-PP butt-joints welded at 174 °C for
various times by tensile test.

Figure 31: SEM micrographs of characteristic failure surface of welded PP/30% (wlo)
GF-PP at 174 °C for 903, (a) at 30% (wlo) GF-PP side. (c) high magniﬁcation
SEM micrograph taken from (a) ((1) high magniﬁcation SEM micrograph taken
from (b)

Figure 32,: SEM micrographs of characteristic failure surface of welded PPI30% (wlo)
GF-PP at 174 °C for 90s, (a) at 30% (wlo) GF-PP side,(b) at PP side,
(c) high magniﬁcation SEM micrograph taken from (a), (c) high
magniﬁcation SEM micrograph taken from (a).

Figure 33: SEM micrographs of characteristic failure surface of welded PP/30% (wlo)
GF-PP at 174 0C for 240 s, (a) at 30% (wlo) GF-PP side, (b) a typical part at

Figure 34:

Figure 35:

Figure 36:

Figure 37:

Figure 38:

Figure 39:

Figure 40:

PP side.

SEM micrographs of characteristic failure surface of welded PP/30%
(wlo) GF-PP at 174 °C for 480 s, (a) at 30% (wlo) GF-PP side,
(b) at PP side, (c) high magniﬁcation SEM micrograph taken from (a).

High magniﬁcation SEM micrographs of failure of interfacial bonding
reformed in weld region, taken from 174 °C, 1203. (a) at a ﬁbre end; (b) Inside
a imprint after a ﬁber debonding.

Schematics for the failure process and surfaces of PP/GF-PP with
shorter welding time

Schematics for the failure of PP/GF-PP with longer welding time

Load-displacement curves for GF-PPIGF-PP butt-joints with various ﬁber
content (wlo) welded at 174 C for 480 s by tensile test

SEM micrographs of characteristic failure surface of welded 30%/30%
(wlo) GF-PP at 174 C for 480 s

Schematics of the failure processes of GF-PP/GF-PP.

CHAPTER 1
INTRODUCTION

Generally speaking, ﬁber reinforced thermoplastics have better mechanical
properties than the unreinforced polymers do, such as higher tensile strength, yield
strength, elastic modulus and relative temperature resistance. Fibre increases the strength
and stiﬂ'ness, but decreases the toughness for some thermoplastics. For example,
polypropylene (PP) is toughner than ﬁbre reinforced polypropylene at room temperature,
however, this is not the case when temperature drops below the glass transition tempe-
rature of PP. It is undoubtable that the increment of toughness in glass ﬁber reinforced
polypropylene (GP-PP) resulted from the intervening of ﬁbres with the matrix. The basic
principle increasing the toughness of resin-based composites lies in the introduction of
various energy-absorbing processes such as ﬁller (particulate or ﬁbre) deformation,
debonding of the interface, ﬁber pull-out, homogeneous and localized plastic deformation
of the matrix, and multiple cracking. The typical theoretical models which have been
proposed to describe the toughening mechanisms including crack pining (Lange, 1970),
crack deﬂection (Farber and Evans, 1983) and crack blunting (Owen, 1979).

The purpose of this study is to investigate the eﬂ‘ect of ﬁber content on the
toughening mechanisms in short glass ﬁber reinforced PP at low temperature. In addition,
the arrangement of ﬁbres in the thickness of injection molded GF-PP plaques affect the
fracture toughness is taken into consideration.

Another separate subject that is the application of GF-PP is the welding of GF-PP
parts both for PP to GF-PP and GF-PP to GF-PP. Short glass ﬁber reinforced PP is widely
used in the industries of automobile and domestic appliances. A good joint is very
important to the integrity of structure as a whole.

The joining of plastics and composites could be dated back to the original study of

amorphous polymers. When two samples of the same amorphous polymer are brought into

contact at a temperature above the glass transition temperature, the interfacial surface
gradually disappear. It is reﬂected by the increase of interfacial strength until. At long
contact times, it is possible to achieve the property of the bulk polymer. At this point the
junction surface has become indistinguishable from any other surface that might be
located with in the bulk material which is described as compeletly “healed” (Voyutskii,
1963). Effect of time, temperature, molecular weight of polymer on welding or healing of
amorphous polymers has been an area of considerable activity, both experimentally (Jud
et al., 1981 and Wool et al. 1982) and theoretically (de Gennes, 1980, Prager et a1. 1980).
Theoretical models have combined the reptation theory of self-diffusion in polymer melts
with molecular criteria and with crossing density criteria, describing an increase in
interfacial strength as interdiffusion progresses to predict the time dependence of the
healing process. However, There is not a successful model to predict the interfacial
strength and fracture energy for welding semi-crystalline polymers and their composites.
Individual study of the strength and interface of semi-crystalline polymers to their
composite joints is very important.

The joint of PP to PP is very easy to approach by various welding methods, such as
hot tool, vibration (Stokes, 1988), but very few reportes on the studies of welding PP to
GF-PP and GF-PP to GF-PP are available in the literature. In this study, the effect of the
processing time and temperature on the strength development of welding PP to short glass
ﬁbre reinforced PP was carried out. The ﬁbres in the weld region involved in the failure of
the welded specimen was taken into consideration. Finally, the mechanical properties of
GF-PP to GF-PP joints were tested and analyzed.

CHAPTER 2

Toughening Mechanisms of Short Glass Fiber Reinforced
Polypropylene

2.1 Introduction

The important parameters determining the mechanical properties of resin-based
composites are the volume fraction of the ﬁller, which may be particulates or ﬁbers, its
particle size or ﬁbre aspect ratio, its modulus and strength, the ﬁber-matrix adhesive, and
the toughness of the matrix.

Short ﬁber reinforced polypropylene has been widely used in automobile and
domestic su'uctures. At room temperature, the matrix toughness plays an important role in
determining the mechanical properties of short ﬁber reinforced polypropylene (GF-PP)
(Gupta et al. 1983). Gupta et al. studied the energy-absorbing mechanisms during fracture
in glass ﬁber reinforced PP and suggested that the total work of fracture included
contributions from the debonding of the interface, the sliding or pull-out of ﬁbers, and the
plastic deformation of matrix. They reckoned that the contribution from matrix plasticity,
either through homogenous deformation of the matrix or from localized deformation
around the ﬁber ends, was dominant.

When the testing temperature drops below glass transition temperature of PP, the
matrix in short ﬁber reinforced PP will become brittle. In this case, the toughening
mechanisms and the main energy-absorbing mechanisms during fracture may differ from
those at room temperature. In addition, injection molding introduces the different
arrangement of ﬁbers across the thickness of molded plaques of short ﬁber reinforced
thermoplastics. This gives rise to various ﬁber-related failure mechanisms (Friedrich,
1989) which has effect on the toughening result.

The objective of present study is to investigate the effect of ﬁber content on the

toughening mechanisms of short glass ﬁber reinforced PP at low temperature.The ﬁber
related failures in GF-PP are determined from the fracture surfaces. The eﬂ'ect of various

arrangement of ﬁbres in GF-PP are analyzed.

2.2. Background

2.2.1. Toughening Mechanisms in Particle Filled Resin-Based Composites

The basic principle increasing the toughness of resin-based composites lies in the
introduction of various energy-absorbing processes such as ﬁller (particulate or ﬁber)
deformation, debonding of the interface, ﬁbre pull-out, homogeneous and localized plastic
deformation of the matrix, and multiple cracking. The typical theoretical models which
have been proposed, including, crack pining, crack deﬂection and crack blunting.

The crack pining mechanism was suggested and modiﬁed by Lange (Lange, 1970),
Lange and Radford (Lange and Radford, 1971) and Evans (Evans, 1972) in the study of
particulate reinforced composites. The increase in toughness of ﬁlled composites was
explained in terms of a crack pining mechanism in that the rigid particles act as obstacles
and force the crack to bow as shown in Figure 1. This model is similar to the interaction
between dislocation and dispersed particles in the plastic deformation of metals. The
interaction between the crack and the reinforced particles occurs during crack propagation
in the case of good ﬁller-matrix adhesion. Therefore, the hindered crack can only bow out
between the well-bonded particles, forming secondary cracks and breaks away from
pining positions when it attains a radius of (DI2), where D is the interparticle spacing. The
toughening effect depends on the interparticle spacing, D, and the effective crack line

tension, T. The maximum toughening effect can be expressed as

G G

T
matrix + 5 (1)

camp =

 

Figure 1: Schematic representation ofthe process ofcrack pinning.

where G and G

comp mm, are the fracture energies of the composite and matrix

respectively. The contribution to T originates from dissipation processes, such as
debonding of the ﬁller-matrix interfaces, local plastic deformation near the interface etc.
Although these results have been obtained primarily through the study of spherical
particle reinforced resin composites, for example, alumina trihydrate epoxy (Lange and
Radford, 1971) and glass beads-epoxy resin (Kinloch, et al. 1989), in principle the, they
may be adequate to describe the behavior of other types of composites ﬁlled by non-
spherical particles (Cantwell et al. 1989). Cantwell et al. showed a linear relationship exits
between the fracture toughness (Kc) and the volume fraction of short glass ﬁbre reinforced
epoxy resins.

The second toughening model is crack deﬂection, which occurs around the second
ﬁller particles. Deﬂection toughening arises whenever interaction between the crack front
and the minor phase produces a non-planar crack, subject to a stress intensity lower than
that experienced by the corresponding planar crack. The non-planar crack arises either
from residual strains present in material or from the existence of weakened interfaces. The
former derive for elastic modulus and/or thermal expansion mismatch between the matrix
and the particulate phase. The sign of the residual strain determines the direction of
deﬂecting (Binns, 1962). Speciﬁcally, a second phase with a greater thermal expansion
coefﬁcient or elastic modulus than that of the matrix, produces tangential tensile strains,
causing the crack to deﬂect toward the particle. When a crack approaches or intercepts a
second phase particle or other kind of ﬁller, it will tilt at an angle, 0, out of its original
plane and subsequently twist at an angle, 4), as shown in Figure 2 (a) and (b) respectively.
The initial tilt angle depends upon the orientation and position of the particle and the
matrix. Farber and Evans (Farber and Evans, 1983) based upon a fracture mechanics
approach to predict the toughness of reinforced composites, both theoretically and experi-
mentally. They pointed out that the increase in toughness depends much upon the shape

and volume fraction of the ﬁller, but is invariant with size of the ﬁller, according to the

 

(b)

Figure 2: Schematics of typical craCk deﬂection: (a) tilt, and (b) twist of
the crack front

mathematical model. The most eﬂ‘ective morphology for deﬂecting propagation cracks is
the rod of high aspect ratio of which short glass ﬁbre is an example. Less effective in
toughening are disc-shaped particles and spheres. The ideal second phase, in addition to
maintaining chemical compatibility, should be present in amounts of 10 to 20 % percent
by volume. Greater amounts may diminish the increase in toughness due to overlapping
particles.

In the case of poor ﬁller-matrix adhesion, the toughening mechanism through
crack tip blunting is important. Evidence of decohesion of the particles ahead of the crack
tip has been given by Owen (Owen, 1979) in a study of the fracture properties of glass
bead-ﬁlled resin. Early debonding leads to blunting of the crack tip, by which the stress
intensity factor, Kc, near the crack tip reduces considerably. It has been observed in some
composites with poor ﬁller-matrix adhesion that the crack tends to propagate in an
unstable stick-slip manner. When unstable propagation occurs, the load-displacement
curve has a characteristic sawtooth shape (Lin and Lai, 1993). It appears that the unstable
stick-slip propagation is probably due to repeated crack initiations and arrests due to the
crack tip blunting.

Except for the above three typical toughening mechanisms in short ﬁber reinforced
composites, ﬁber pull-out and deformation of matrix and ﬁbers are very important energy
consuming mechanisms which intervene during crack propagations (Gupta et al., 1990).
However, it is very difﬁcult to predict their contribution to the total fracture energy of ﬁber
pull-out by qualitative study because it is a dynamic process during the failure of com-
posite (Cantwell et al. 1989). The crack pinning, crack deﬂection, crack blunting and other
energy absorbing mechanisms such as ﬁber pull-out deformation of matrix and ﬁbers,
debonding of ﬁller/matrix could occur simultaneously or separately during fracture of
reinforced composites by particulates and ﬁbers or other kinds of ﬁllers (Faber and Evans,
1983)

2.2.2 Linear Elastic Fracture Mechanics
The subject of fracture mechanics was developed from the basic concept of
Grifﬁth (Griffth, 1920) dealing with brittle materials. The strength of a brittle material is

governed by small defects that act as stress concentrators, in accordance with the equation:

1

2 E 5
a, = (-137) (2)

where:
a, - tensile strength
E - tensile modulus
y - fracture surface energy

c - the length of intrinsic defect

Figure 3 is a typical diagram presenting a realistic crack in an inﬁnite specimen to

calculate the crack opening mode stress intensity factor. The result was concluded as:
K 1 cc Yen/1; (3)

and the strain energy release rate for the opening mode is

K?

G, = f (4)

It is assumed that a crack will propagate when K , reaches a critical value, K ,c,
which is speciﬁc for each material. G ,c, is respectively, is the critical strain energy release
rate. Kc, the critical value of stress ﬁeld intensity factor, is derived from the tensile
strength of a sharply notched specimen by a transformation based on the specimen

geometry. It is a relatively simple physical quantity and the related G c is even simpler and

10

HH’HH
MW

iii

 

 

M
titiitit

Figure 3: A center-cracked panel of inﬁnite plate loaded at the remote
boundaries.

11

both quantities are considered to be the main measures of toughness. The simple expre‘
ssion for Kc, at the tip of length 2a, in the centre of an inﬁnitely wide sheet subjected to a

tensile stress acting in the direction perpendicular to the axis of the crack, is placed by:

Kc = elk/Z (5)

Y - the geometry correlation factor.
For diﬂ'erent testing method and specimen, eq.5 can be modiﬁed according to the
geometry correlation factor. The fracture toughness and geometry correlation can be

expressedin the following formulas (Turner, 1983):

Fmax a
Kc = WNW) .,/2 (6)

and

14%) = 29.6-185.5(%)+655.7(%)2-1017(%)3+633-9(j£y)4 (7)

where:

Fm - the maximum tensile load
B - the thickness of specimen
W -the specimen width

a . a
(W) should satisfy W s 0.5

Thermoplastics and some of the short ﬁbre reinforced thennOplastics are not brittle
at room temperature, and they exhibit extensive plastic deformation. Moreover, these
materials are not crack sensitivity at room temperature. Fraster and Ward (Fraster and

Ward, 1977) applied a fracture mechanics analysis to results obtained on specimens of

12

polycarbonate bearing notches with various tip radii and concluded that only in the case of
razor-notched specimens is it possible, with complete certainly, to interpret the data
directly in terms of fracture toughness. Specimens with blunter notches give self-

consistent results and higher values of fracture toughness.

2.3. Experimental Procedures

2.3.1. Determination of ﬁber length distribution and arrangement

Chopped glass ﬁber reinforced PP with 5,10, 20 and 30% percent ﬁber by weight
(wlo) and neat PP plaques, produced by injection molding were received from Himont
Inc, Lansing, Michigan. The dimensions of the molded plaques were 101.6 mm x 25.4
mm x 3.2 mm. Fibre length or aspect ratio and the arrangement of ﬁbers in the short
ﬁber reinforced composites are very important parameters to aﬁ'ect the mechanical
properties of the material. It was necessary to frist determine the ﬁbre length and ﬁber
arrangement in the thickness before analysis of the GF-PP fracture properties could be
perfomed.

The process was begun by cutting a small piece of each composite and putting
them in separate clean aluminum pans, then heating them in an open oven at about 400 °C
until the resin was burned off completely. The next step invoved gently placing the ﬁber
remaining from each of the corresponding composites into water gently then depositing
them onto ﬁlter papers.The ﬁbre length, up to a total number of 200, were then measured
under a microscope. The ﬁber length distribution and the average length for each
composite was determined.

Injection molding of short glass ﬁber reinforced PP often resultes in a natural
laminate or layer structure across the thickness. To observe the ﬁber arrangement across
the thickness, (the shaded plane), a small piece of each specimen at the center of the strip

was cut from each composite as shown in Figure 4. These samples were polished to diﬂent

l3

 

 

 

 

 

 

 

 

 

 

101.6 mm
. . Z [I 3.2mm
Injectron 4 Y
25.4
K t - - mm
X

Figure 4: Technique for cutting samples to detect ﬁber orientation in injection
molded short GF-PP plaque

14

depth and observed under an optical microscope.
2.3.2. Fracture Toughness Testing

The values of fracture toughness of short glass ﬁber reinforced PP were deter-
mined using the sharp notched compact tension conﬁguration according to ASTM
(D5045-1992). Location of CT specimen with respect to injection direction and its dimen-
sions are shown in Figure 5. (a) and (b). The sharp notch was created by tapping the
sample with a liquid nitrogen cooled blade. Two types of precrack, longitudinal (L-notch)
and transverse (T-notch) to the injection direction, were made in the CT specimen in order
to taking into account the eﬂ'ect of the ﬁber arrangement in the thickness direction on the
fracture toughness. The fracture tests were conducted on a XI-6 series Instron mechanical
testing machine under displacement control. All compact tension tests were performed at a
constant crosshead speed of 2.54 mrnlmin, and at a temperature of about ~87°C achieved
by irnbedding the CT specimen into dry ice in a chamber. The temperature was recorded
by a thermocouple and the ﬂuctuation of the temperature during the test was within :l:5 °C.

The fracture toughness in terms of the critical value of the stress ﬁeld intensity
factor, K c, was calculated using equation 6. Each result represents the average of ﬁve
tests. After testing, the fracture surfaces of the specimens were coated with gold by

vacuum evaporation and examined by a Scanning Electron Microscope (SEM).

2.4. Experimental Results and Discussion

2.4.1 Fiber length Distribution in GF-PP

Figure 6 shows the distribution of ﬁbre length for the short glass ﬁber reinforced
pp with various ﬁber content. The average ﬁber lengths, '1, in 5%, 10%, 20% and 30%
(wlo) are about 0.85 mm, 0.80 mm, 0.70 mm and 0.64 mm respectively. The diameter of
the glass ﬁber is about 14 um. The ﬁber length decreases slightly with increasing ﬁber

volume fraction, which causes ﬁber-ﬁber interaction during the mixing of ﬁber and matrix

15

 

 

Injection

 

 

 

 

 

 

 

 

 

 

 

\
. ’ 1er
1 B=3.2 mm

‘ W=20 mm .

 

 

 

 

(b)

Figure5: (a) shows how to cut Cl‘ specimens from the received injection
molded PP and GF-PP plaque and (b) the geometry and dimen-
sions of CT-specimen according to ASTM.

16

 

(a)

Percent ("/o)

 

 

 

 

 

 

 

 

 

 

 

      

 

   

 

7 :j‘? _'
(b) .. --
0 —- N v- \o- co '— ‘T °°.
Fibre length (mm)
a; 40
a. 30
g; 20
(C) ‘3 10
A 30
.\° 23
j 20 «
8 104 9'
(d) a I 4 3 1
a. 0 J f. 11.2; .. . . 7..
"‘ N. ‘1 V‘. ‘0. 5. °°. "‘ N. ‘0
o o o o o o o —- --
Fibre Length (mm)

 

 

Figure 6: Illustrates the distribution of ﬁbre length (mm) in the injection
molded short GF-PP by number percent: (a) for 5% (wlo); (b) for
10% (wlo); (c) for 20% (w/o) and (d) for 30% (w/o) GF-PP.

17

and the injection molding process. The difference in the ﬁber length in GF-PP with
various ﬁber content is not a factor affecting the fracture toughness. This can be concluded
from the result achieving the maximum toughening by rod-shaped particle when the
aspect ratio is above 12 (Faber and Evans, 1983). The aspect ratios of most the ﬁbers are

above this value, which can be determined from the distribution of the ﬁbers in Figure 6.

2.4.2. Fibre arrangement across the thickness of the molded plaques

Figure 7 (a), (b), (c) and (d) show microscopically the arrangement of the ﬁbers
across thickness (Y 2 plane) of injection molded short glass ﬁber reinforced PP with 5%,
10%, 20% and 30%(w/o) respectively. The small circles represent those ﬁbers arranged
along the injection direction (X-axis), and the ellipsoids represent those ﬁbers misoriented
along X-axis. It could be found that those ﬁbers in the center region, (C), occupy about
0.90 mm, which is one quarter to one third of thickness of a molded plaque. At both side
regions, (S), most ﬁbers are arranged randomly. This sandwich-laminate structure in the
thickness is obvious, especially in the plaques with ﬁbre content higher than 10% (wlo).
Thus, the injection molded short glass ﬁber reinforced PP can be featured as a layer-
structure in which there is a center region, (C), where most ﬁbers are arranged along the
injection direction, and in which at both side region, (S), the ﬁbres are random in the
thickness direction. This layer-structure is symmetric about the center as shown in Fig. 8,

schematically.

2.4.3. Toughening Mechanisms in Short GF—PP

Notched compact tension specimens of PP and short GF-PP were tested by static
tension load at dry ice temperature (-87 °C). The typical deformation of load-opening
displacement are illustrated in Figure 9, with curves-l, 2 representing PP and 30% (wlo)
with T-notch respectively. The deformation curves of 5%, 10% and 20% (wlo) are

between the curves of PP and 30% (wlo). From the deformation curves, the fracture

18

(thickness)

(width)
(injection)

 

(a)
Figure 7: (a) Fiber arrangement in Y2 plane of 5% (wlo) GF-PP (XSO)

19

(thickness)

(wrdth)

(injection)

 

(b)
Figure 7: (b) Fiber arrangement in Y2 plane of 10% (wlo) GF—PP (X50)

AZ (thickness)

 

—’(v3t{dtrt)

(injection)

 

 

(c)
Figure 7: (c) Fiber arrangement in Y2 plane of 20% (wlo) GF-PP (X50)

21

(thickness)

(width)

(injection)

 

(d)
Figure 7: (d) Fiber arrangement in Y2 plane of 30% (w/o) GF-PP (X50)

 

 

 

 

 

 

 

S
C
I ‘ c
Injectiondirec " S
9 O
.. ~0 2.. .XT of:
_= O

 

I -.<-. ’-

Figure 8: Injection molded induced layer structure in short GF-PP plaque.
S indicates side region and C-center region, schematically.

23

 

 

150
_ 2
100
52‘
'93 1
.—J
50 _.
1 r I n
0.08 0.10

 

Displacement (in)

Figure 9: The deformation of notched CT specimens of PP and GF-PP under
static load by load-openin g displacement, 1 for PP, 2 for 30% (w/o)

GF-PP with T-notch.

24

behavior of notched (71' specimens of PP and GF-PP can be recognized approximately as a
linear brittle fracture.
Table 1 lists the fracture toughness in terms of the critical stress intensity factor,

K c, and data of neat PP to 30% wlo GF-PP with L-notch and T-notch.

Tablel. Fracture Toughness, Kc, of PP and GF-PP

 

 

 

 

 

 

 

 

 

 

specimen Ifc (MPa e mm) standard deviation
Neat PP 4.23 0.15
5% W/O L-notch 4.97 0.22
5% wlo T-notch 5.30 0.14
10% wlo L-notch 5.81 0.17
10% wlo T-notch 5.93 0.25
20% W/O L-notch 5.97 0.29
20% wlo T-notch 6.36 0.20
30% wlo L-notch 6.26 0.29
30% wlo T-notch 6.63 0.22

 

 

 

Figure 10 is the plot of fracture toughness with the ﬁber content for L-notch and
T-notch.This illustrates that the value of K c increases fast from PP to 10% (wlo), then
goes up slowly from 10% to 30% (wlo), both for L-notch and T-notch CT specimens. The
incremental differences in the fracture toughness of GF-PP, in accordance with to the ﬁber
content, indicate that the toughening mechanisms might be varied with the ﬁber contents.

The toughening mechanisms of GF-PP and the mechanism of energy-absorping in

pure PP are discussed as follows:

2.4.3.1. Mechanism of Energy-absorption in PP

The value of fracture toughness of notched CI‘ Specimen of PP is 4.23 MPa 0 my2

25

.0"
m \r

O‘
I

 

CI ----for T-notch

 

 

 

 

Fracture Toughness, Kc, (MPaﬁ)
{It
in

 

5 " I ----for T-notch
4.5 y
4 . t t t t 3
0 5 10 15 20 25 30
Fibre Content (w/o %)

Figure 10: Fracture toughness, Kc, of notched CT specimens of PP and
short GF-PP under static load and at low temperature (-87 0C).

26

at the testing conditions. This value is quite high compared with other polymeric materials
(Kausch, 1978). This indicates that the matrix itself has good crack resistance at low
temperature.

Figure 11 is a SEM micrograph representation the fracture surface of PP near the
notch root, which shows a dimple pattern of deformed matrix in this region. It must be
accepted that the dimple pattern of deformed polymer matrix is characteristic of crazing in
the amorphous and semi-cystalline polymers (Friedrich, 1983). In the study of polymer
fracture, it has been concluded that crazing plays an important role in the failure of PP at a
range of temperature from -200 °C to 150 °C (Kitagawa et al. 1984). The formation of the
dimple pattem in this fracture surface can be explained from the spherulite structure of
semi-crystalline PP. The spherulites are pointed, nucleated semi-crystalline entities which
grow in a spherically symmetrical form until their boundaries impinge. The ﬁbrils formed
fmm the amorphous part near the boundaries of the spherulites under normal load. Inside
the spherulite, is a crystal structure, therefore, the plastic deformation is conﬁned to the
weak spherulites boundaries and results in a typical dimple fracture pattern. The failure of
fribril, either by continued chain reparation including slippage and disentanglement of
molecular coils, or through scission of entangled chains (Hannon, 1974). Friedrich
(Friedrich, 1983) concluded that under plane strain condition (favored by sharp notch, low
temperature), crazing provided a mechanism for absorption and dissipation of fracture
energy.This explains pure PP could retain a relatively high fracture toughness, even at a

temperature below its glass transition temperature.

2.4.3.2 Low ﬁbre content of GF-PP

Figure 10. shows that the fracture toughness increases quickly from 5% to 10% (w/
o) and the relationship of Kc to the ﬁber content is approximately linear (for example, the
plot of T-notch specimens). This phenomenon can be explained as a crack pinning

mechanism, in which the rigid short glass ﬁbers act as obstacles and force the cracks to

 

Figure 11: SEM micrograph of fracture surface of notched CT specimen of
PP near the notch root showing dimple pattern of the deformed
matrix.

28

how between them. Figure 12 and Figure 13. represent the fracture surfaces near the
notches of 5% and 10% (wlo) GF-PP with T-notch respectively. The existence of “tails”
behind the ﬁbers on the fracture surface has been cited as evidence of crack pinning
mechanism in particle reinforced resins (Kinloch and Young, 1983). Those “tails” were
formed behind the ﬁbers (pull-out ﬁbers or bmken ﬁbers) as a result of the interaction
between the propagating cracks and the ﬁbers after the crack front had broken away from
the latter. The fracture surfaces appear ﬂat except for the localized matrix deformation
around the ﬁbers and homogeneous deformation in the large ﬁber avoid areas.

From the above observations, it is believed that crack pinning can be used to
explain the toughening mechanism of short glass ﬁber reinforced PP at low temperature.
The increase of fracture energy is proportional to the inverse of the interparticle spacing
(D). The best approximation for the interparticle spacing between randomly arranged
rod-shaped particles (short ﬁbres) is that given by Bansal and Ardell (Bansal and Ardell,
1972), where the center-to-center nearest neighbor spacing between ﬁnite parallel
cylinders

4v "
3- if I .

2 4V,
Vf

NI—

e"dx (7)

where Vf is the volume fraction of ﬁber, and d is the diameter of ﬁber. D can be deter-
mined in the GF-PP as about 95 um, 54pm,39|.tm, and 28 pm for 5%w/o(2% vol), 10%
wlo(4% vol), 20% wlo(8% vol) and 30% wlo(l3% vol) respectively. According to the
results in studies of short carbon ﬁber reinforced polyphenylene sulphide (PPS) and
phenolphthalein side-group polyethersulphone (PES-C) thermoplastics by Lin and Lai
(Lin and Lai, 1993), and the reports in other cracking pinning systems (Lang and Radford,
1971), it was found that the fracture energy increased linearly with D varying from inﬁnite
to about 40pm. Smaller D revealed little effect on the increment of the fracture energy.

However when the spacing is larger than 40 pm, the crack is “ﬂexible” allowing the crack

 

Figure 12: SEM micrograph of fracture surface of notched CT specimen of
5% (w/o) GF-PP with T-notch near the notch root.

30

 

Figure 13: SEM micrograph of fracture surface of notched CT specimen of
10% (wlo) GF—PP with T—notch near the notch root.

31

front to bow out between the particles, thereby forming secondary cracks under the action
of the driving force. Thus, energy is required not only to create the new fracture surface
but is also supplied to the newly formed non-linear crack front, resulting in an increase of
fracture energy or fracture toughness. When the value of D decreases further, the
ﬂexibility of the crack reduces and the crack pinning mechanism is no longer valid for
predicting the fracture toughness any longer. Thus, the crack pinning in GF-PP with ﬁber

content below 10% (wlo) fractured at low temperature is the main toughening mechanism.

2.4.3.3 High Fibre Content of GF-PP

The crack pinning and crack deﬂection processes usually occur simultaneously.
Crack pinning produces a non-linear crack, and crack deﬂection produces a non-planar
crack. Figure 14 and Figure 15 represent the fracture surfaces near the notch roots of 20%
(wlo) and 30% (wlo) with T-notch respectively, which display non-planar fracture
surfaces. Although post-failure examination of fracture surfaces does not allow determina-
tion of the precise path and direction of the crack (or a trace of the crack front) at any
instant during propagation, it is capable of providing valuable information about the crack
deﬂection proﬁle (Faber and Evans, 1983). Thus, the existence of crack deﬂection in the
fracture process of 20% and 30% (wlo) GF-PP is possible and reasonable.

The plot of the mathematical model, based on a fracture mechanics approach, is
shown in Figure 16 for three aspect ratios (Fraber and Evans, 1983). It predicts that the
toughening increment becomes a volume fraction invariant above ~0.2. The inﬂuence of
the aspect ratio tends to be an asymptotic limit at larger values of 12. The fracture tough-
ness of notched CT specimens of GF-PP slowly increment from 10% wlo (4% vol) to 30%
wlo (13% vol) in this expriment is consistent with the result predicted from Figure 16,
which shows in the range of ﬁber volume fraction from about 0.05 to 0.2.. Greater amount
of ﬁbre content may diminish the toughness increase due to overlapping particles. The

phenomenon of overlapping ﬁbres in the fracture surfaces are shown clearly in Figure 14

32

 

Figure 14: SEM micrograph of fracture surface of notched Cl‘ specimen of
20% (wlo) GF-PP with T-notch near the notch root.

33

 

Figure 15: SEM micrograph of fracture surface of notched CT specimen of
30% (wlo) GF-PP with T-notch near the notch root.

34

00 h
l l
a
K.

.Relottve toughness .47 ‘m
N

 

 

1 l . r
' .
0 CI 02 0.3 04
Volume frocrion of rod , V,

Figure 16: Relative toughness predictions from crack deﬂection model for rod-
shaped particles of three aspect ratios (Faber and Evans, 1983).

35

and Figure 15.

In addition to the crack pinning and crack deﬂection mechanisms of toughening,
ﬁber pull-out, homogeneous and localized plastic deformation of the matrix, debonding of
the ﬁber-matrix interfaces and a few ﬁbers broken, are also very important absorbing
energy mechanisms in the fracture of short glass ﬁber reinforced PP. The evidence for all
of these ﬁber-related failures can be observed from the fracture surfaces both in the low

ﬁber and high ﬁber content specimens (Figure 12 to 15.)

2.4.4 The Effect of Fiber Arrangement on The Fracture Toughness

For the same ﬁber content of CT specimen, a T-notch has a higher value of fracture
toughness than that of a L-notch specimen. This can be explained through the layer struc-
ture in the thickness of injection molded short glass ﬁber reinforced PP plaques. The ﬁber
arrangement in the T-notch and L—notch can be illustrated schematically as Figure 17 (a)
and (b), respectively. In the notched plane of a CT specimen with a T-notch, there is a
center region, (C), in which about one-fourth to one-third of those ﬁbers are arranged
transverse to the notch direction, and the ﬁbers are random in the both side regions, (S).
On the other hand, in a L-notch specimen, those ﬁbers in the center region are longitu-
dially arranged along the notch direction, and ﬁbers are random at both sides. In this case,
there are a relatively larger number of ﬁbers which are arranged nearly perpendicular to
the advancing crack front in the T—notch samples. The transversely arranged ﬁbers can
contribute much more eﬂ'ectively to the strength of the material in front of the starter
crack, and require higher loads to force those cracks to move forward. If, on the other
hand, more ﬁbers are located in plane , almost parallel to the crack direction, the crack
could propagate along the ﬁber sides by void formation and cause ﬁbre/matrix debonding
which has less eﬂ‘ective crack hindrance eﬂ‘ect (Friedrich, 1989). Figure 18 represents the
characteristic fracture surface of 20% (wlo) GF-PP with a L-notch showing the area in the

center region, (C). Comparing Figure 18 to Figure 14 (20% wlo with T-notch and taken

36

from the center region of the thickness) it can be observed clearly, that more ﬁbers are
arranged longitudially to the notch root in 20% (wlo) CI‘ specimen with L-notch, however
more ﬁbers are arranged transverse to the notch direction in the latter. This explains the
reason why the value of fracture toughness of the specimen with T-notch is higher than
that of the with L-notch.

37

 

 

 

 

 

 

 

 

 

 

 

 

 

(b)

Figure 17: Schematics to show ﬁbre arrangement in T-notch (a) and
L-notch (b).

38

 

Figure 18: Characteristic fracture surface of 20% (wlo) GF-PP with L-
notch shows large number of ﬁbres were arranged parallel to the
notch.

2.5 Summary
The increment of fracture toughness, is dependent on the ﬁber content at low

temperature. When ﬁber content is low (from 5% to 10% wlo), Kc increases rapidly and
the mechanism of toughening can be explained by crack pinning. However, the fracture
toughness increases at a slower rate when ﬁber content is above 10% (wlo), and the
toughening mechanism explained by crack deﬂection. In addition, other energy
consuming mechanisms intervene during crack propagation and exist simultaneously,
such as ﬁber pull-out, plastic deformation of matrix (homogeneously and localized),
debonding of the ﬁber-matrix interfaces and a few ﬁbers broken. Under current testing
conditions, crazing provided a mechanism for absorption and dissipation of fracture
energy for notched specimen of PP.

It was found that the CT specimen with T-notch has a higher value of fracture
toughness than that with L-notch for the same ﬁber content. This is because more ﬁbers
were arranged transverse to the prenotch in the T-notch specimen than that in the L-notch

specimen according to the layer structure of injection molded GF-PP.

39

CHAPTER 3
Welding of PP and Short Glass Fiber Reinforced PP

3.1 Introduction

Welding or healing of amorphous polymers was studied widely both experimen—
tally (Jud et al., 1981 and Wool et al. 1982) and theoretically (de Gennes, 1980, Kim and
Wool, 1981, Prager et al. 1980) based on the reptation theory which was developed by de
Gennes (de Gennes, 1971). De Gennes considered the movement of a single chain in a
ﬁxed network, which although an ideal case, gives very good results conformed by exper-
iments for amorphous thermoplastics.

The reptation theory predicts that the length of the chain ends (or called minor
chains) varies with the square root time. Later, wool and Kim combined the reptation
theory of self-diﬂ‘usion in polymer melt with molecular criteria, which include molecular
weight, the number of chains, number of bridges, average monomer interpenetration
depth, etc., describing the relationship of healing processing time and the macrosc0pic
mechanical properties, such as interfacial strength and fracture toughness. Their theoreti-
cal model states that the energy of fracture is proportional to the number of chain segments
spanning the interface, and obtained the fracture stress is proportional to the fourth root of
time. Those results are in agreement with the experiments of healing amorphous polymers
conducted by Jud etc., (Jud et al., 1981, and Wool and O’Connor, 1982).

The study of welding thermoplastic composites was carried out by Bastein and
Gillespie (Bastein and Gillespie, 1991). They used the reptation model to predict the
fracture stress and energy as a function of the process history including temperature and
time. However, their model is restricted only to the amorphous thermoplastic composites,
because they used an amorphous ﬁlm bonding technique, by which placed a thin ﬁlm of
amorphous material in the interface.

Although there have been some investigations into the welding of semi-crystalline

4o

41

polymers and their composites by Stokes (Stokes, 1988, 1991) and Andrew and Bevis
(Andrew and Bevis, 1984), they concentrated in the ﬁelds of developing welding methods
and the effect of processing parameters on the strength of butt joints. In these experimental
studies, the welding temperature were chosen above the melting temperatures of the
matrix and thus, the welding processes depended upon the molten ﬂow. The interface
between the two parts disappeared at the instance of keeping the two parts intimate
contact. There is not a successful model to predict the interfacial strength and fracture
energy for welding semi-crystalline polymers and their composites. The investigation of
the relationship of the interfacial strength of semi-crystalline polymers and their compo—
site joints is very important.

In this study, the effect of the processing time and temperature on the interfacial
strength development of welding PP to short glass ﬁber reinforced PP and GF-PP to itself
was carried out. The ﬁbers in the weld region involved in the failure of the welded

specimen was taken into consideration.

3.2 Background

3.2.1 Strengthening Mechanisms of Short Fibre Reinforced Thermoplastics

In the case of continuous ﬁbre reinforcement, all the ﬁbers were working at
maximum efﬁciency, with the average strain in the ﬁber being equal to that in the matrix.
Assuming that bonding of the interface between ﬁber and matrix is perfect, the tensile

modulus in the tensile direction; modulus of the composite can be estimated as equation 8,
Br: = Emtbm‘lrEﬂf (8)

and the modulus, Ed, in the transverse direction can be estimated by equation 9.

42

1_¢m 9f 9
E—c;--E—;+E; ()

where £05.. and E, are tensile moduli of the composite, matrix and ﬁber respec-
tively. ¢m and (it; are the volume fractions of the matrix and the ﬁbers in the composites.
It is more difﬁcult to deal with the short ﬁber reinforced composites due to the
complication of a spectrum of ﬁbre length and orientations caused by manufacturing
processes, such as injection molding and extrusion. An earlier study on the strength of
short ﬁber reinforced thermoplastics was carried out by Cox (Cox, 1952). The calculation
of the variations in the shear and tensile stress along a short elastic ﬁbre in a plastic matrix,
was estimated by elastic mechanics. In aligned short ﬁbre composite, Figure 19, shows the
state of deformation around a short ﬁbre. The ﬁbre restricts the deformation of the
surrounding matrix, because it is stiffer than the matrix material. The load is transferred
from the matrix to the ﬁber throught the interfacial shear stresses. The shear stress is
greatest at the end of the ﬁber, and decays to are some where along it’s length. The
tensile stress is zero at each end of the ﬁber and reaches a maximum at the center. If the
ﬁber is at a length equal or greater than the critical elastic aspectic ratio, (1%) c, the
maximum tensile stress reaches the tensile stress of the ﬁber. Figure 20. summarizes the

results.

3.2.2 Fracture mechanisms of Semi-crystalline polymer

The problem of deformation behavior of semi-crystalline polymers becomes more
complicated when a crystalline-amorphous microstructure of the polymer must be
considered. Brady and Yhe (Brady and Yhe, 1981) showed that the crystalline spherulites
deformed only after substantial deformation had already occurred in the interconnecting

amorphous regions. Crazing and shear-banding were the primary modes of the deforma-

tensile stress

shear stress

43

 

 

 

 

 

 

lllll'll'll lSlll/{l
_ 1: - I

(a) Before applying load (b) Under tensile load

I

 

 

Figure 19: The effect of a ﬁbre on the matrix deformation.

 

 

 

 

 

 

\ K

 

 

 

 

 

 

 

 

 

 

L

> (5)c

u
{I}
DI t~

< (gt -f§< (g) g

C

Ulh

Figure 20: The tensile stress and shear stress distribution along a ﬁbre.

 

44

tion mechanisms for semi-crystalline polymers.

The formation of crazes in semi-cystalline thermoplastics is a common response to
an applied tensile stress. Under plane strain conditions, (favored by thick sections, sharp
notches or at low temperatures), crazing provides a mechanism for absorption and dissipa-
tion of energy. In the past, craze-like features have been frequently observed during the
studies of semi-crystalline polymers, although they were sometimes named differently,
such as “strain bands” in ﬁbres of PA, PETP, PE and PP, “cracks” and “microcrack” in PP
and PA 6, in certain literatures (01f and Peterlin, 1974). The craze formation in bulk semi-
crystalline thermoplastics has been concluded by Friedrich (Friedrich, 1983) as the
following three steps:

I. First of all, at sites of enhanced stress concentration, (i.e at preexisting voids in
the microstructure or at sites where amorphous layers between lamellae are oriented
nearly under 45. to the applied load), a local sliding occurs between individual lamellar
ribbons having only a low degree of entanglrnent in their common amorphous interlayer.
At this stage, strain is accommodated almost entirely by the interlamellar amorphous
regions. During this process the crystals change their orientation, and the shear stress
component decreases, while the stress in direction of the molecular axis of the crystals
increases.

11. In a secondary step, once the local tensile stress reaches a critical value (equal
to the stress at which macroscopic yielding can take place), individual blocks are pulled
out of the crystal ribbons. Due to this local yielding process, submicroscopic defects,
having an ellipsoidal shape, are created between the lamellae. However, the defects are
able to cause an increase in stress in their lateral neighborhood, consequently involving
adjacent lamellar ribbons in the local deformation process. Thus, the probability for
further creation of voids, as well as for formation of ﬁbrils between these voids, is
increased.

III. The third step is the longitudinal transmission of strain onto connected crystal-

 

 

45

line blocks and leads to a perfect stretching of these ﬁbrils. As the ﬁbrils are able to
stabilize the enhanced microvoid volume between them, a lateral coalescence of these
voids ﬁnally provides a local deformation zone in the shape of a craze, as known from
amorphous polymers.

Figure 21. is a schematic diagram for these three steps.

The formation of shear bands of semi-crystalline polymers during deformation has
been found only when it was loaded by compression and at the temperature lower than
0.75 T: (Friedrich, 1983). Few comprehensive studies exist which consider the formation

of shear bands for semi-crystalline polymers.

3.2.3 Fracture Mechanisms of Short Fiber Reinforced Thermoplastics

Short ﬁber in thermoplastics can transfer the load from matrix to ﬁber and result in
greater strength, as well as other mechanical properties. The mechanisms of the fracture of
the composite are complicated because they are greater aﬂ‘ected by the ﬁber length, the
ﬁber orientation and the ﬁber/matrix bond strength.

It seems that Sato et al. (Sato et al., 1984) studied the mechanism of fracture of
short ﬁber reinforced thermoplastics comprehensively. In their study, short glass ﬁbrer
reinforced polyarnide 6.6 was investigated by in situ SEM. It was found that there were
three stages within the failure. Under a tensile load, the stress concentration occurs in the
vicinity of ﬁbers, especially at the ﬁber ends. Under the inﬂuence of the concentration
stress, cracks may occur at the ﬁber ends. This is the ﬁrst stage of failure. While the shear
deformation of the matrix along the ﬁber sides increases with stress, the interfacial cracks
propagate from the ﬁber ends along the ﬁber sides due to the shear deformation of the
matrix in the second stage of the failure. With the occurrence of the these cracks, the load
bearing capability of the ﬁbers is greatly reduced, and the matrix is required to support a
greater load. The matrix cracks come to grow from the interfacial cracks just prior to the

composite failure. These interfacial cracks and matrix cracks may result in a catastrophic

 

 

+ stress

.33.
l

33

amorphous interlayer

/

33333.0 %%

\\30

as
m
b

:\O
33 33%

. 3

 

Figure 21: Schematic steps of crazes formation in a semi-cystalline polymer structure

47

crack propagation leading to the composite failure. This is the ﬁnal stage of the failure.
Even in this stage, ﬁbers are only pulled out from the matrix and without broken. This
mechanism of the composite fracture may explain the brittle manner, in contrast to the
ductility of the unﬁlled polymers. The results can be diagrammed in Fig. 22. (a), (b), and
(c).

3.2.4. Welding of Thermoplastics and Composites

Welding can be achieved by using various methods for plastics. All of these
methods can be classiﬁed in terms of the heat generations employed, such as hot gas, hot
plate, and infrared welding, etc. In spite of the diﬂ‘erences between these welding tech-
niques, the fundamental of all fusion bonding processes is nothing but intermolecular
diffusion between surfaces in intimate contact (Bastien and Gillespie,l991).

The earliest systematic study of healing or welding in amorphous polymers was
undertaken by Voyuskii (Voyutskii, 1963), who gave the explanation of the molecular
interpretion as shown in Figure 23. Before contact, the situation is shown in (a), with the
polymer chains being prevented by energetic considerations from extending isolated
tendrils through the polymer-air surface. This is also the case immediately after contact,
before signiﬁcant conﬁgurational relaxation has had a chance to occur, as shown in (b).
Although the density of polymer segments is now the same as in the bulk near the junction
surface, the conﬁgurational distribution of the polymer molecules certainly is not; since
each chain is still required to lie entirely to one side of the junction. In time, of course,
conﬁgurational relaxation does occur, leading to an increasing number of intersections of
polymer chains within the junction surface (c), which is reﬂected macroscopically in the
growing mechanical strength of the joint.

In last decades, healing of cracks, in samples of glassy polymers was studied
comprehensively (Jud and Kausch, 1979; Jud et. al, 1981; Wool and O’Connor, 1981).
Theoretical models have combined the reptation theory of self diffusion in polymer melts

(Gennes, 1971) with molecular criteria describing an increase in interfacial strength

F

(a) First stage; occurrence of crack at ﬁbre ends

(b) Second stage; propagation of the crack along the ﬁbre sides

 

(c) Final stage; occurrence of the crack in the matrix

Figure 22: Model of fracture mechanism of short ﬁber reinforced thermoplastics.

49

 

 

 

 

 

 

@233 $33

(a) Before contact of two polymer specimen

 

 

 

 

 

 

 

(b) After contact, the polymer chain conﬁgurations
near two surfaces start to change

@%%

(c) After enough contact time, polymer chains relaxiation

 

 

occur and the interface disappear

Figure 23: Molecular model of healing in polymers.

50

interdiffusion progresses to predict the time dependence of the healing process.Prager and
Trrrell (Prager and Trrrell, 1981) studied the crossing density criterion, which states that
the energy of fracture is proportional to the number of chain segments spanning the inter-
face, and obtained an initial J; dependence of the fracture energy in agreement with
experimental data. This corresponds to a ‘1/3 of the growth of fracture stress.

The literature on the welding reinforced thermoplastics is quite small. They are
mainly concerned the welding methods (Andrews and Bevis, 1984). Andrews and Bevis
investigated the butt-fusion welding of chopped-graphite-ﬁbre-ﬁlled polyvinydeneﬂuo-
ride(PVDF) to itself and to the neat material. In their study, it was shown that the weld
strength of the ﬁlled material was signiﬁcantly lower than the strength of the ﬁlled
material. The strength of ﬁlled to unﬁlled material joints could, however, be almost as
high as that of the unﬁlled material. The reduction of the strength in the weld joint of com-
posite to composite was considered owing principally to the off-axis alignment of ﬁbre in
the weld plain due to the melt ﬂow in the welding process. Strokes (Stokes, 1991) studied
the welding of chopped-glass modiﬁed polyphenylene oxide (GF-MPPO) to itself and to
the neat material by vibration-welded butt joints with results similar to those involving the
joints of PVDF and ﬁlled PVDF. Besides off-axis alignment, poor bonding at the ﬁbre-

resin interface in the weld zone, is another reason for the strength reduction of the joint.

3.3. Experimental Procedures

3.3.1. Material

The materials which are used to make butt-joints are injection molded PP and 5%,
10%, 20% and 30% wlo GF-PP. In addition to these materials, some ASTM tensile speci-
mens of PP and 5% to 30% wlo GF—PP were also received from Himont company, which

can be used as a reference for the study of strength development.

 

3.3.2 Welding Processing

A welding tool was designed as shown in Figure 25. The heat, generated from the
hot plate is transferred through the aluminum tool. The on-axis pressure loading can be
applied driving the screw bolt, which moves the aluminum block in the groove. The
pressure is controlled by the displacement of scew bolt, in which one tum is equal to 51:
inches. This design will provide an even force distribution on the surface of the welding
specimen; moreover, it can limit the ﬂow of the melting materials, which produces a weld
bead outside the weld region in most other welding methods. In this case, the eﬂ‘ect of
weld bead on the mechanical properties can be overlooked. The temperature was
measured by a thermocouple, and temperature ﬂuctuation was controlled within i2 °C.
The techniques for welding, PP/GF-PP and GF-PP/GF-PP butt-joints are as follows:

(1) The injection molded GF-PP plaque is cut as in Figure 5 (a), to a 26 mm length,
with the weld faces perpendicular to the injection direction. The faces to be welded are
then polished and parallel.

(2) The whole surfaces of the specimen are cleaned with acetone and dried .

(3) The welding tool is placed in the hot press and the temperature is checked by a
thermocouple through a makeup-specimen. When the temperature goes up to the required
temperature and equilibriums, two parts of specimen are placed inside the groove and
covered . The two parts are then kept in contact by driving the scew bolt, at which point
rimming of the process begins.

(4) After the faces of the two parts have been held together under a chosen time,
temperature, and pressure, the weld tool is taken from the hot press and cooled down by
placing it between two large aluminum blocks. At the cooling rate of about 56 °Clmin, it
only need 10 seconds are need to cool the specimen from 180 °C to 165 °C (at or below
this temperature, no joint can be made). The thermal history of the welding process is
shown in Figure 26. Due to the short period of time involved with cooling, its eﬂ'ect on the

51

52

 
 
     

hot press
_
. §\\‘\\\\\\_\}\\\\\§ Wﬂ/ﬂ Li S K\\\\\\\\\\\\\\‘. 4—scrcw bolt

hot press

two parts of specimen aluminum block

“P

I

down

 

 

 

welding tool

thermocou I la

Figure 25: Schematic of the welding tool and equipment.

53

weld joint is not signiﬁcant.
(5) When a visually satisfactory weld bar had been produced, the edges and the
surfaces, which have thermallyl contracted in the cooling stage, are polished. The dimen-

sions of the ﬁnal weld bar is 50 mm x 12 mm x 3.0 mm.

Temperature

 

 

 

 

 

to tp Time

Figure 26: Thermal history of welding process.

3.3.3. Tensile tests
Tensile tests of ASTM tensile specimens and the welded butt-joints of PP/PP,
PPIGF-PP, and GF-PP/GF-PP were carried out on a XI-6 series Instron mechanical testing

machine at a constant cross head speed of 2.54 rum/min and at room temperature.

3.4. Experimental Results and Discussion

3.4.1. Tensile Properties of Bulk PP and GF-PP

The ASTM (D638-89) tensile specimens of PP and short GF-PP were tapered at j:
the center, with a diameter, d, of 0.08 inches. This was done because preliminary testing of
dumb-bell shaped specimen of PP and 5% wlo GF-PP always resulted in failure at one of
the grips or shoulders. Therefore, in order to insure that the failure occurs in the range of
the gauge region, tapered is necessary, and this method has been used in the tensile test for
soft plastics. The tapered hole in the specimen does not eﬂ‘ect the peak and yield strength,
but eliminates the break strain signiﬁcantly. Figure 27 represents the load-displacement
deformation curves of the injection molded PP and short GF-PP tensile bars, and curves-1,
2, 3, 4, and 5 represent PP, 5%, 10%, 20%, and 20% (wlo) GF-PP specimens respectively.

The results of tapered ASTM tensile specimens are summarized in Table 2.

Table 2:. The results of ASTM tensile specimen

 

 

 

 

 

 

specimen No. Peak stress Yield stress Young’s Break strain
(MPa) (MPa) modulus (GPa) (96)
PP 35:1: 0.3 22.3 :1: 1.5 0.80 :l: 0.05 13.0 :1: 0.8
5% GF-PP 39 :l: 1.0 27.0 :t 2.0 0.97 :l: 0.03 7.2 :1: 0.6
10% GF-PP 44 :1: 2.4 33.6 :1: 2.5 1.18 :l: 0.1 5.6 :1: 0.4
20% GF-PP 51 :l: 1.6 42.2 :1: 3.5 1.52 :i: 0.06 4.5 :l: 0.3
30% GF—PP 56 :l: 1.6 48.2 :1: 2.7 1.84 :i: 0.06 4.1 :l: 0.3

 

 

 

 

 

 

 

 

500

400

300

Load (lb)

200

100

55

H ——---—-.—-—

 

I I l l I 4
0.05 0.100 0.150 0.200 0,250 0.300

Displacement (in)

Figure 27: Load-displacement curves for injection molded ASTM tensile
bars of PP and short GF-PP with various ﬁbre content (w/o), 1, 2, 3,
4, and 5 present PP, 5%, 10% 20%, an 30% GF—PP, respectively.

56

In Table 2, it is shown that the peak and yield stresses of composites increase steadily with
ﬁbre content. The peak and yield stresses increase about 80% and 100% from PP to 30%
wlo GF-PP respectively. Young’s modulus increase from 0.8 GPa for PP to 1.84 GPa for
30% wlo GF-PP. However, the break strain decreased abruptly, from PP to 5% wlo GF—PP,
and then gradually dropped off with increased ﬁbre content.

Figure 28. represents the typical failure surface of a 30% wlo GF-PP ASTM tensile
specimen. From this picture, it can be observed that a large number of holes and outwardly
protruding ﬁbres on the failure surface indicates that ﬁbres were pulled out completely
during the failure of composite. The mechanism of this failure can be explained by the
theory developed by Sato et al. (Sato et. al. 1984). Under a tensile load, the stress concen-
tration occurrs in the vicinity of the ﬁbres, especially at the ﬁbre ends. When stress
increased, the shear stress increased and it was greatest at the ﬁbre end, as shown in Figure
20. Under the inﬂuence of stress concentration and the shear stress, the cracks were
resulted at the ﬁbre ends and propagated along the ﬁbre sides, forming interfacial cracks.
When the peak stress is reached, the interfacial cracks propagated into the matrix and

caused the ﬁnal composite failure.

3.4.2. Welding Parameters on Strength Development

The butt-joints of PP/30% (wlo) GF-PP were welded according to the techniques
described in section 3.3.2 and tested as described in section 3. 3.3. Table 3 records the
peak stress of the welded bars of PP/30% (wlo) GF-PP at different welding times and
temperatures. The welding pressure was controlled in terms of the displacement by
driving the screw bolt. The displacement for welding all the butt-joints is maintained
constant by keeping the two parts contact when they were just placed into the weld tool..

Figure 29. is the plot of peak stress of welded PPI30% (wlo) GF-PP bars with
various welding times and temperatures by tensile test. From these curves, the time
needed to reach a plateau strength of PP/30% (wlo) GF-PP can be determined by the

\ m.

 

 

57

various welding time and temperature

Table 3: The peak stress of welded joints of PPI30%GFRe-PP with

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

T1=167 °C T2=l70 °C T3=174 °C T4=180 °C
t1=30 s no no no 23.5 d: 2.4
t2=60 3 no no no 29.2 :I: 0.3
t3=90 s no no 24.7 :i: 1.5 31.5 :1: 0.4
t4=120 s no 16.3 :1: 3.7 30.0 d: 1.0 30.4 :1: 0.8
t5=150 s no 21.2 :I: 1.8 29.0 :1: 2.0 I
t5=180 s 15 :l: 0.8 26.5 :1: 2.8 32.0 i 0.3 31.3 :t 0.3
t7=210 s I 31.5 i 1.2 I l
t3=240 s 23.5il.4 32.2 i 1.4 32.9 :1: 0.8 I
t9=270 s 29.8 :1: 1.1 I I I
t10=300 s 31.5 :1: 1.6 30.5 :1: 1.6 I 31.3 :1: 0.6
tu=360 s 32 :l: 1.4 I l l
t12=390 s I 30.4 :1: 1.6 I /
t13=480 s 30.0 :1: 1.7 30.4 :I: 1.3 31.8 :1: 1.3 31.4 :1: 0.2
Note:
“no” means no joint at all.

“I” means not available

the unite ofpeak stress is MPa

 

58

 

Figure 28: SEM micrograph of characteristic failure surface of short
GF-PP 30% (wlo) of ASTM tensile test, viewed down on
the surface.

 

59

 

 

 

 

 

5000 I I I I
/ . -----------
0' ,x’“ V ,_ i if _______
O\ '0 ,Z—_____ 5 r. ,__ 8
o /,.-" '- v _. _________________ v ......... II".:.\..,,._ v
:' v V
o ,«o
(73 4000 —
Q4 s7
m f, .5
g1) .
E O 1 7 4 C v
m l 8 0 C
5d 5:
:5 $7
0.. 3 0 0 0 —- :
f {1670
V i
1 '7 O C
2 000 1 l J 1
0 2 4 6 8

WELDING TIME (min.)

Figure 29: Plot of peak stress of welded PP /30% (w/o) GF-PP bars with
various welding times and temperatures by tensile test

60

corresponding to the welding temperature. The approximate time for reaching the ultimate
strength is about 270 s for 167 °C, 210 s for 170 °C, 120 s for 174 °C and only 60 seconds
for 180 °C. In the experiment, the butt-joint can not be achieved even after 16 minutes, at
the welding temperature of 165 0C, beyond this time, PP will degrade. The higher the
welding temperature, the shorter the time for reaching the ultimate strength for a butt-
joint. When the welding temperature approaches the melting temperature of the matrix
(178 °C). the time needed to heal the joint is much shorter. This is because the process of
healing a polymer joint is more rapid through melting ﬂow than by polymer diffusion
alone. These welding parameters can be used to predict the welding of PP and its compo-
sites in industry.

The load-displacement curves of welded PP/30% (wlo) GF-PP at 174 °C for
various times under tensile test are illustrated in ﬁgure 30., whereby curve-1, 2, 3, and 4
present the welded bar, welded for 903, 1205, 2408 and 480s, respectively. Although the
strength of PPI30%(w/o) GF-PP has reached the ultimate strength when the welding time
is longer than 1208 at 174 °C, the elongation of the butt-joints increases with the welding
time. From the data in Table 3, the ultimate tensile strength of PPI30% (wlo) GF-PP is
about 32 MPa, which is very close to the peak strength of PP, 35 MPa, if the experimental
error is taken into consideration. The developed strength of this butt-joint is much lower
than that of 30% wlo GF-PP bulk material, 56 MPa. This result is consistent with the study
of welding PVDF to it’s chopped-graphite-ﬁbre ﬁlled PVDF, which has shown that the
strength of ﬁlled to unﬁlled material joints could be almost as high as that of the unﬁlled
part (Andrews and Bevis, 1984).

The ultimate strength of a weld bar of PPI30% (wlo) GF-PP is almost the sameas
that that of bulk PP. This is not surprising, because when a weak material is jointed to a
stronger material, even if the joint is perfect, the strength of such a joint specimen can only
reach the strength of the weaker one. The most interesting observation is that failure

surfaces of the joint specimens change the appearance with the welding time as well as the

 

61

 

 

200 — .
, 3
2.
g l 1 4
“C3
3
.J
100 ——
I I l I l l I J
0.01 0.02 0.03 0.04 0.05 0.06 0.07 ' 0.08

Displacement (in)

Figure 30: Load-displacement curves for PP/30% GF-PP butt-joints
welded at 174 °C for various times by tensile test, curves-1, 2, 3,
and 4 represent for 90 s, 120 s, 240 s, and 480 s respectively.

62

elongation.

Scanning Electron Microscopy inspection of the specimens, after testing, indicated
that the failure occurred at or near the weld region (or weld junction), which is related to
welding time.

Figure 31. represents the failure surface of PPI30% (wlo) GF-PP welded at 174 °C
for 90 s, (a) at reinforced side: (b) the matched part at PP side; (c) and (d) showing higher
magniﬁcations of a small part of (a) and (b) revealing the ﬁbres ends on the surface of
reinforced PP and corresponding imprints left on PP side, respectively. From (a) or (b), it
is illustrates that the failure of the this butt-joint occurred in the weld surface or junction.
The peak strength of this specimen is 24.7 MPa, about 75 percent of the ultimate strength,
32 MPa. The pictures of (c) and (d) show detail of the failure surface at the reinforced and
PP sides. The ﬁbres have not protruded outwards into the PP side, and the imprints left on
PP side are very shallow and irregular, which indicates that the bond between the ﬁbre
ends and matrix did not form in the weld surface, explaining why the strength is lower
than the ultimate strength.

When the welding time was increased to 120 s, the peak strength reached the
ultimate strength, but most failure area were in the weld junction, as shown in Figure 32.,
(a) at reinforced side; (b) at PP side. Figure 32. (c) and (d) are viewed at a higher magniﬁ-
cation to detail the typical appearance of (a) and (b), respectively. It can be clearly seen the
ﬁbres had protruded outwards from reinforced part and into the PP side by the evidence of
clearly imprints left in the PP side, as evidenced by the after the failure of joint. However,
those outwardly protruding ﬁbres are very short therefore they are most likely located in
the same plane which acted as stress concentrators in the whole specimen, under tensile
load.

Figure 33 represents the failure surface of PPI30% (wlo) GF-PP welded at 174 °C
for 240 s, (a) at reinforced part; (b) the typical surface appearance on the PP side. The

failure surface still coincides with the weld junction, because there are some protruding

63

 

(a)

Figure 31: SEM micrographs of characteristic failure surface of welded
PPI30% (wlo) GF-PP at 174 0C for 903, (a) at 30% (wlo) GF-PP
side.

 

(b)
Figure 31 --Contd. (b) at PP side

 

(C)

Figure 31 --Contd. (c) high magniﬁcation SEM micrograph taken
from (a)

 

(d)

Figure 31 --Contd. (d) high magniﬁcation SEM micrograph taken
from (b)

67

 

(a)

Figure 32.: SEM micrographs of characteristic failure surface of welded
PPI30% (wlo) GF-PP at 174 °C for 90s, (a) at 30% (wlo) GF-PP
side.

 

(b)
Figure 32 ---Contd. (b) at PP side

69

 

(C)

Figure 32 ---Contd. (c) high magniﬁcation SEM micrograph taken
from (a)

70

 

(d)

Figure 32 --- Contd. ((1) high magniﬁcation SEM micrograph
taken from (a)

71

 

,\
(a)
Figure 33: SEM micrographs of characteristic failure surface of welded
PPI30% (wlo) GF-PP at 174 °C for 240 s. (a) at 30% (wlo) GF-
PP side.

72

 

(b)
Figure 33 n-Contd. (b) A typical part at PP side.

73

ﬁbres on the reinforced part, and correspondingly, some holes and imprints can be
observed on the PP side. The obvious difference in Figure 33 as compared to the prior two
ﬁgures, is the relatively longer of the protruging ﬁbres. The appearance of the failure
surface at the reinforced part is obviously different form the failure surface of bulk 30%
(wlo) GF-PP tensile specimen, as shown in Figure 28, where the outwardly protruding
ﬁbres in the failure plane are much longer, and some resin remaining on them which
indicates the reforming bonding in the weld region is better than the specimen welded for
shorter time.

When the welding time increases at the same temperature, 174 °C, most of the
failure surface of butt-joint is out of the weld region. Figure 34 is the failure appearance of
butt-joint of PPI30% (wlo) GF-PP welded at 174 °C for 480 s, (a) at reinforced part across
the thickness; (b) partial area at the PP part and (c) higher magniﬁcation of a small part of
(a) to show the detail of transition from ductile to brittle deformation. The whole appear-
ance of the failure surface is convex-concave. There are only some pull-out ﬁbres in the
center region which occupy about 8 mm (along the wide of weld bar) in the length and 0.6
mm in width (along thickness), in addition, the deformation of matrix in this region is
ductile, but the rest of the area around the center ductile region is brittle failure, and
occurred in the PP side. This is known because there is not any debris of ﬁbres can be
found, as shown in (a) or (b). In the region of ductile deformation, is the feature of ﬁbril
coalescence near the pull-out ﬁbres, which resulted in the rough bundles of deformed
matrix. This convex-concave appearance can probably be explained through the layer
structure of the injection molded GF-PP, as studied in the last part, and the movement of
ﬁbres during the process of welding. The layer structure of the injection molded short
GF-PP existed along the thickness, where in the center region, (C), most ﬁbres were
arranged longitudially but at both side regions, (8), the ﬁbres were random. In the welding
process, due to diﬁ‘ering ﬁbre concentrations at both sides (reinforced part 30%(on), PP

side it is 0%), is different at both sides and the concentration gradient could drive the ﬁbre

74

 

(a)
Figure 34: SEM micrographs of characteristic failure surface of welded PPI30%
(wlo) GF-PP at 174 °C for 480 s, (a) at 30% (wlo) GF-PP side.

75

 

(b)
Figure 34 -—-Contd. (b) at PP side

76

 

(C)

Figure 34 ---Contd. (c) high magniﬁcation SEM micrograph taken
from (a)

77

towards the PP side. The ﬁber orientation on the weld surface has an important eﬁect on
the movement of the ﬁbres, whereby the ﬁbres in the center region are orthonormal to the
weld surface and parallel to the movement direction, and therefore, need a lower driving
force. However, those ﬁbres in the side regions are random, some lying ﬁbres, perpendi-
cular to the move direction, are difﬁcult to move forwards, or move much more slowly
than those ﬁbres in the center. For longer welding times, the ﬁbres in the center region will
move longer distances into the PP side, but the ﬁbres at both side regions move very little
or stand still at the weld surface. In addition, a longer welding time could beneﬁt for the
reformed bonding of ﬁbrelmatrix in the weld region.

The differences in failure surface of PPI30% (wlo) GF-PP butt-joints with varying
welding times can be better understood with the following explanation . When the welding
time is short, but longer enough to reach the ultimate strength of the joint, the ﬁbre ends in
the reinforced part lie just at the surface, or just across the junction. Those ﬁbre ends or the
lying ﬁbres are concentrated on a single plane -weld junction. When the welded bar is
loaded, the crazes will form just below the ﬁbre ends or at ﬁbre sides, as shown in Figure
35 (a) the fribril coalescence near the ﬁbre end and (b) the crazes formed at the left side of
the lying ﬁbre and growth to the right along the ﬁbre/matrix interface. These crazes grow
and form microcracks in the weld junction. Due to the high concentration of ﬁbre ends in
the single plane, those microcracks approach each other gradually. When they coalesce to
some crack size, rapid propagation of the failure will occur in the weld junction, and
resulted in a ﬂat surface. However, when the welding time is longer, the ﬁbres in the center
region (C) will move a longer distance than those ﬁbres at both side regions, (S), and
result in the ﬁbres being found in the center layer bridge in the reinforced part and PP part.
Under tensile load, high stress concentration will occur in this area and the cracks form
through the ﬁbrils coalescence and broken. When the crack reaches a critical size, the ﬁnal
failure will occur in the PP side instead of the weld junction, because of lower energy

through this path, and result in a failure surface with a convex-concave appearance.

78

 

(a)

Figure 35: High magniﬁcation SEM micrographs of failure of inter-
facial bonding reformed in weld region, taken from
174 °C, 1203. (a) at a ﬁbre end;

79

 

(b)
Figure 35 ---Cont. (b) Inside a imprint after a ﬁber debonding.

m
8
.m
D.
.m

on PP side

 

 

 

crazes forming just under ﬁbre

(a) under load

atics for the failure process and surfaces of PP/GF-PP with
ends; (b) the failure occurring in the weld region.

shorter welding time,

Figure 36: Schem

81

 

    
  

|.
VIII

   
 

   

      
 

\‘ \ \‘
.t §§~S
§§§§\
V“ F‘e“
\‘
. \\“

  

o uctile deformation

brittle deformation

 

Figure 37: Schematics for the failure of PP/GF-PP with longer welding time,
(a) For a longer welding time, the ﬁbres in the center region go
further into PP part than ﬁbres in both side regions. crazes forming
ﬁrst in the center region; (b) ﬁnal failure through PP part and
appearance being convex-concave.

82

The schematic diagrams for the failure processes and failure appearance of
PPI30% (wlo) GF-PP, welded for shorter and longer times are shown in Fig. 36. and
Fig. 37. respectively.

3.4.3 Tensile Properties Butt-Joints of GF-PPIGF-PP

The butt-joints of 5%I5% w/o, 10%/10% wlo, 20%/20% wlo and 30%/30% wlo
GF-PP were welded at 174 °C for 480 seconds. The deformational behavior under tensile
tests of the welded bars are illustrated in Figure 38, which are load-displacement curves of
5%/5%, 10%/10%, 20%/20% and 30%/30% wlo GF-PP from left to right. The results of

tensile tests are summarized in Table 4.

Table 4: Results of tensile test of GF-PP/GF-PP butt-joints

 

 

 

 

.. 222:2 22.22 .2222. 2:2
5%!5% wlo 32 :1: 1.4 20.4 :I: 0.9 0.58 :1: 0.02 16.5 :i: 1.3
10%/10% wlo 35 i 1.0 27.7 :1: 2.1 0.65 :1: 0.04 10.0 :1: 0.7
20%/20% wlo 32 :1: 0.3 30.6 i 1.1 0.76 :1: 0.03 8.0 :1: 0.6
30%/30% wlo 33 :1: 0.8 31.5 :1: 1.7 0.71 i 0.05 7.3 :1: 0.4

 

 

 

 

 

 

 

From Table 4 and Figure 38, it can be concluded that:

(i) the peak strength reached by those butt-joints is almost the same regardless of
the ﬁbre content of the two weld parts.

(ii) yield strength and Young’s modulus increases with the ﬁbre content of the two
weld parts, and the break strain decreases with ﬁbre content.

Visual inspection of the welded butt-joints of GF-PP/GF—PP after tensile testing
indicated that all the failure occurred at the weld plane, even for times greater than 480 s,
but the mechanical properties were almost the same as those welded for 240s and 480s.

83

This indicates that the weld region is always the weakest place in the whole butt-joints of
GF—PPIGF-PP. For example, Figure 39 represents the failure surface of 30%/30% (wlo)
GF-PP welded at 174 °C for 4803 by SEM, (a) at one side failure specimen; and (b) higher
magniﬁcation of a small part of (a). From (a), it can be observed that the failure occurred
in the weld region. The deformation of this surface is dominated by the feature of matrix
as compared to the failure surface of 30% (wlo) GF-PP tensile specimen, as shown in
Figure 28. In addition, only a few imprints have been left on this surface, which were
supposedly to be made by the ﬁbre ends of another part. Those two phenomena might hint
to the existance of a thin layer of matrix existed between the two weld surfaces of 30%
(wlo) GF-PP, which is possible if taking the polished surfaces of both weld parts into
consideration. Most the ﬁbre ends on the polished surface became broken due to polish-
ing, and result in a very uneven surface at the ﬁbre ends as shown in Figure 39 (b). When
the two weld surfaces are brought into contact, a gap is formed between any two Opposite
ﬁbres and provides the space in which a thin layer of matrix can form. In addition, the
same ﬁbre content for both weld parts will resist the forward ﬁbre movement and result in
fewer ﬁbres penetrating the weld region, and bridging the two parts of the weld specimen.
There is also the another reason for the existence of a thin layer of matrix between the two
polished surfaces. The existance of a thin layer of matrix in the weld region with very few
ﬁbres able to bridge the two parts, are the main reasons for the weld region being the
weakest place in the whole specimen of GF-PP/GF-PP, as well as the strength of GF-PPI
GF—PP being limited to the strength of bulk PP regardless of the ﬁbre content in the two
weld parts. It certainly results the failure occurring in the weld region of any butt-joints of
GF—PPIGF-PP.

The failure process of butt-joints of GF-PPIGF-PP can be explained by the follow-
ing. When the welded specimen is loaded by tensile load, a cluster of crazes will form near
the ﬁbre ends in the weld region since this is the weakest place. Those crazes could grow

down and up and approach each other in the thin ﬁlm of matrix in the weld region. They

 

o
o—
m
/4 2
3 .
O
o_ 1
N
A
.D
C
'0
53
O
.J
O
S—

 

I
! I

' ' ' 007
0.01 0.02 0.03 0.04 0.05 0.06 .

Displacement (in)

Figure 38: Load-displacement curves for GF—PP/GF-PP butt-joints with
various ﬁbre content (w/o) welded at 174 C for 480 s by tensile

test: 1 for S%/5%; 2 for 10%/10%; 3 for 20%/20%; 4 for 30%/30%
welded bars.

 

(a)

Figure 39: SEM micrographs of characteristic failure surface of welded 30%/30%
(wlo) GF-PP at 174 C for 480 s, (a) at one side of the failed specimen.

86

 

(b)
Figure 39 ---Contd. High magniﬁcation SEM micrograph taken from (a).

87

coalesce and form voids or cracks near the ﬁbre ends, and result in the ﬁbre ends separat-
ing from the matrix. When the cracks reach a critical size, the brittle failure occurs in the
weld region and causes the ﬁnal failure of the whole specimen. All of these failure
processes are illustrated schematically by Figure 40, (a) butt-joint of GF—PP/GF—PP; (b)
crazes forming at the ﬁbre ends in the weld region; and (c) cracks propagating in the thin
layer of matrix, resulting the ﬁnal failure of the joint in the weld region.

Unlike the strength of GF-PP/GF-PP being limited to the strength of bulk PP, the
yield strength and modulus increase with the ﬁbre content in the two weld parts, as
recorded in Table 4. Both of them are not only relative with the properties of the weld
region, but also relative with the properties of the both sides. The both side of weld region
are bulk reinforced PP, which are much stiffer than the material in the weld region because
there are no effectively reinforced ﬁbres and a thin layer of matrix exists in the weld
region. The deformation will occur in the weld region ﬁrst under loading and then
restricted in this region by both sides. The higher the ﬁbre content in the two weld parts,
the stronger the restriction of the both sides acting upon the welding region. However, as
the difference between the property of the weld region and the weld parts at both sides
becomes greater and greater, the property of weld region will have more important eﬁ‘ect
on the whole butt-joint. This is why the yield strength and modulus of the butt-joints of
GF—PPIGF-PP increase with ﬁbre content in GF-PP from 5% to 20% (wlo), but do not
change much from 20% to 30% (wlo), as shown in Table 4. 0n the other hand, the break

strain of this kind of joint will decrease with the ﬁbre content in the two weld parts.

w\\\\>..

crazes forming

at ﬁber end

mwamw

Alva; .

.474...

i
.\\\\\\b ..

H r.
.

 

mm

.mg

mm

cm mm
.8 .U 6
mm WMM
tau m m
«w rma
mu mma
m m 3.

   

           

     

\\\\\ wow/Z
II5.

A O
”222%?
7////A :2K\\\

7////0

.ennrr

  

propagation in the thin layer of matrix and
result the ﬁnal failure in the weld region

 

(c) crazes coalescence to forming cracks and

(a) butt-joint of GF-PP/GF-PP

Figure 40: Schematics of the failure processes of GF-PP/GF-PP.

3.5 Summary

A systematic study of the butt-joints of PP to short GF-PP and GF-PP to GF-PP
has been carried out.

The welding time required to reach the ultimate strength of PF] 30% (wlo) FG-PP
was determined under various temperatures. The minimum temperature for PPI30% wlo
GF—PP was found to be 167 °C. The higher the temperature, the shorter the time needed to
obtain the ultimate strength for the butt-joints. The ultimate strength for the butt-joints of
PP/GF-PP is close to the peak strength of bulk PP, which is the weaker one in the welded
bar.

The failure appearance of PP/GF-PP is affected by the welding time, even after the
ultimate strength has been reached. For a shorter welding time, the failure place always
coincides with the welding surface, owing to most of the ﬁber ends and lying ﬁbers from
the reinforced parts being located on a single plane. For a longer welding time, the failure
starts at the center area of the weld region with ﬁnal failure through the PP part, resulting
in a convex-concave appearance. This is because most of the ﬁbers in the center region are
arranged along the injection direction, and the ﬁbers at both side regions are random
according to the layer structure of injection molded GF-PP. The ﬁbers in the center region
can move easier than those in the side regions and are able to move further into the PP
part. The stress concentration occurs near those ﬁber ends and generates cracks through
crazes forming at this area, then causes the ﬁnal failure in the PP side.

The reinforcement of ﬁber in PP is cancelled out by introducing a weld into the
system of GF-PPIGF-PP. The peak strength this joint can reach limited to that of bulk PP,
because there are very few ﬁbers able to penetrate the weld plane and bridge two weld
parts, with a thin layer of matrix existing in the weld region. Owing to the diﬁ'erent prop-
erties between the weld region and the two weld parts, the stiﬂ‘er material at both sides of

weld region will restrict the occurrence of deformation and propagation in the weld

89

90

region, and the failure of the butt-joint of GF-PPIGF-PP will result at the weld region,
which is the weakest place in the whole bar.

CHAPTER 4
CONCLUSIONS

The increment of fracture toughness, Kc, is dependent upon the ﬁber content at
low temperature. When ﬁber content is low (from 5% to 10% wlo), Kc increases quickly
and the mechanism of toughening can be explained by crack pinning. The fracture
toughness increases much slower when ﬁber content is above 10% (wlo) and in this case,
crack deﬂection is best to be used to explain the toughening mechanism. In addition, other
energy consuming mechanisms intervene during crack propagation and exist
simultaneously, such as ﬁber pull-out, plastic deformation of matrix (homogeneously and
localized), debonding of the ﬁber-matrix interfaces and a few broken ﬁbers.

Under current testing conditions, crazing provides a mechanism for absorping and
dissipation of fracture energy for a notched specimen of PP.

It was found that the Cl‘ specimen with the T-notch had a higher value of fracture
toughness than that with the L-notch for the same ﬁber content. This is because more
ﬁbers were arranged transverse to the prenotch in the T-notch specimen than that in the

L-notch specimen, according to the layer structure of injection molded GF-PP.

The second part investigated the butt-joint of PP to short GF-PP and GF-PP to
itself.

The welding parameters of time and temperature on the strength development of
PPIGF-PP was carried out and it was found that the higher the temperature, the shorter the
time needed to obtain the ultimate strength for the butt-joint. The minimum temperature
for PPI30% wlo GF-PP is found to be 167 °C. The ultimate strength for the butt-joints of
PPIGF-PP is close to the peak strength of bulk PP, which is the weaker part one in the
welded bar.

The failure appearance Of PPIGF-PP was found to be eﬁ'ected by the welding time,

91

92

even after the ultimate strength has been reached. For a shorter welding time, the failure
occurred in the weld region, owing to most of the ﬁber ends and lying ﬁbers from the
reinforced parts being located on a single plane. For a longer welding time, the failure
starts at the center region of the weld region, then ﬁnal failure through the PP part, and
results in a convex-concave appearance. This is because most of the ﬁbers in the center
region are arranged along the injection direction and the ﬁbres at both side regions are
random according to the layer structure of the injection molded GF-PP. The ﬁbers in the
center region can move easier than those in the side regions and move further into the PP
part in a relatively longer time. The stress concentration occurs near those ﬁber ends and
generate cracks through the formation crazes of at this area and then cause the ﬁnal failure
in the PP side.

The reinforcement of ﬁber in PP is cancelled out by introducing a weld into the
system of GF-PP/GF-PP. The strength that this joint can reach is limited to that of bulk PP
since there are very few ﬁbers able to penetrate the weld plane and bridge two weld parts,
with a thin layer Of matrix existing in the weld region. There is the main reason for the
reductional properties of the butt-joints of GF-PP/GF—PP. The reinforced PP at both sides
of weld region is stiffer, restricting the deformation in the weld region, and the failure
place of the butt-joint of GF-PP/GF-PP will result at the weld region, which is the weakest

place in the whole bar.

BIBLIOGRAPHY

Gupta, V.B., Mittal, R.K., and Goel, Malti: Compos. Sci. Technol. 37 353 (1991)
Friedrich, K. and Karger-Kocsis, I .: Fractography and Failure Mechanism of Polymers
and Composites, Elsevier Science Publishers LTD, London and New York, 437-494
(1989)

Lange, F. F.: Phil. Mag. 22 938 (1970)

Lange, F. F. and Radford, K. C.: J. Mater. Sci. 6 1197 (1971)

Evans, A. 6.: Phil. Mag. 26 1327 (1972)

Kinloch, A. 1., Maxwell, D. A. and Young, R. 1.: J. Mater: Sci. Lett. 4 1276 (1987)
Cantwell W. I. and Anne C. Roulin-Moloney: F ractography and Failure Mechanisms of
Polymers and Composites, edited by Anne C. Roulin-Moloney, Elsevier applied Science ,
London and New York, 233-290 (1989)

Binns, D. 3.: Science of Ceramics, edited by Steward, G.H., Academic Press, New York.
315 (1962)

Faber, K. T. and Evans, A. 0.: Acta Metall. 31 S65 (1983)

Owen, A. B.: J. Mater: Sci. 14 2523 (1979)

Lin, G. M. and Lai, J. K. L: J. Mater: Sci. 28 5240 (1993)

Griffth, A. A.: Phil. Trans. Roy. Soc. 221 163 (1921)

Turner, 8.: Mechnical testing of Plastic, 2nd edition, (1983)

Fraser, 1: A. w. and Ward, 1. M.: .L'Mater: Sci., 12 459 (1977),

Baneal P. P. and Ardell, A. 1.: Metallography, 5 92 (1972)

Kausch, H. H.: Polymer Fracture, Berlin-Heidelberg-New York springer, (1978)
Harmon, MJ.: J. Appl. Polym. Sci. 18 3761 (1974)

Kitagawa, M. and Yarnmaura, H.: J. Mater: Sci. 19 1863-1875 (1984)

Jud, K., Kausch, H. H., and Williams, J. G.: J. Mater: Sci. 16 204 (1981)

93

94

de Gennes, P. G. de: J. Chem. Phys. 55 572 (1971)

Wool, R. P. and O’Connor, K. M.: J. Polym. Sci. Polym. Letters Ed., (1981)
Wool, R. P. and O’Connor, K. M.: J. Poly. Sci., 20 7 (1982)

Kim, Y. H. and Wool, R. P.: Macromolecules, 16 115 (1983)

Prager, S., Adolf, D. and Trrrell., M.: J. Chem. Phys. 78 7105 (1983)
Storkes, V. K.: Polym. Eng. Sci. 28 728 (1988)

Strokes, V. K.: Polym. Eng. Sci.: 31 511 (1991)

Battien L1. and Gillespie, J. W. 1R.: Polym. Eng. Sci., 31 1720 (1991)
Andrews, 1.R.F. and Bevis, MJ. 1.: Mater: Sci. 19 645 (1984)

Cox, H. L.: J. Appy. Phys.: 3 72 (1952)

Brady, T. E., Yeh, G. S. Y.: J. Mater: Sci., 5 573 (1970)

Olf,H.G. A., Peterlin, H.G.A.: J. Polym. Sci., Polym. Phys. 12 2209 (1974)
Friedrich, K.: Advanced in Polymer Science, Vol. 52I53 (1983)

Sato, N., Kurauchi, T., Sato, S. and Kamigaito, O.: J. Mater: Sci., 19 1145 (1984)

Voyutskii, S. S.: Autohesion and Arﬂresion of High Polymers, Wiley-interscience, New
York. (1963)

Ind, K. and Kaush, H. H. : Polymer Bull. l 697 (1979)

nrcurcnu STATE UNIV. LIBRARIES
ill1|WWIIWIWNINIWINIll”WWIWWI
31293010258980

 

.592;