Composites: Part A 37 (2006) 1796–1804

www.elsevier.com/locate/compositesa

High-tenacity man-made cellulose fibre reinforced thermoplastics

– Injection moulding compounds with polypropylene and
alternative matrices
J. Ganster *, H.-P. Fink, M. Pinnow
Fraunhofer Institute for Applied Polymer Science, Geiselbergstr. 69, 14476 Potsdam-Golm, Germany

Received 8 June 2005; received in revised form 8 September 2005; accepted 8 September 2005

Abstract

High-tenacity man-made cellulose filament yarn (rayon tyre cord yarn) has been used to reinforce polypropylene (PP), polyethylene,
high impact polystyrene (HIPS), and poly (lactic acid) (PLA) for injection moulding applications. Highly homogeneous composites are
obtained with the pultrusion compounding method developed. For PP the influence of coupling agent, fibre weight fraction, fibre cut
length and PP type on basic mechanical properties of the composites have been studied. For a fibre load of 30 wt%, typical values
for tensile strength, modulus, Charpy unnotched and notched impact strength are 80 MPa, 3.5 GPa, 85 kJ/m2, and 12 kJ/m2, respec-
tively. A high impact resistance level is maintained also at low temperatures where the matrix material becomes brittle. For the other
matrix materials, similar reinforcing effects are observed, except for the impact behaviour of HIPS, where the reinforcing fibres interfere
with the impact modification of the matrix polymer. In contrast, the impact characteristics of PLA are drastically improved increasing
the unnotched and notched Charpy strengths by 380% and 200%, respectively. With the property level obtained, cellulose man-made
fibre reinforced composites prove to be an alternative to short glass fibre reinforced plastics.
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: A. Polymer–matrix composites; A. Thermoplastic resin; D. Mechanical testing; E. Injection moulding

1. Introduction both in terms of tensile strength and, in particular, impact

strength the cellulose fibre reinforced polypropylene com-
In the past decade, the reinforcing potential of cellulosic posites do not reach the (short) glass fibre level (see, e.g.,
and lignocellulosic fibres in thermoplastics, first of all poly- [6] for random mats or [13,14] vs. [15] for injection mould-
propylene, has attracted much interest in the literature (see, ing compounds). Apart from the lacking uniformity of nat-
e.g., reviews [1,2]). Besides wood fibres, paper, and pulp ural products, one of the main problems with plant fibres
(e.g., [3–5]), special attention has been paid to annual plant seems to be the internal composite structure of the lignocel-
fibres [6], such as, e.g., flax [7], jute [8–10] or sisal [11]. lulosic fibre. In order to take full advantage of the favour-
Compared to glass fibres, as the predominating reinforcing able intrinsic mechanical properties of the cellulose I
component for polypropylene [12], cellulosic fibres offer a crystals, the individual cellulose fibrils, which are fairly uni-
series of advantages: lower density (1.5 g/cm3 vs. 2.5 g/ form and have the desired high aspect ratio needed for effi-
cm3), better recyclability and disposal, reduced abrasion cient reinforcement, must be separated. This poses
to processing machinery, and CO2 neutrality [6]. However, considerable practical problems and has not been achieved
yet. Moreover, the low elongation at break of both the fi-
*
Corresponding author. Tel.: +49 0 331 568 1706; fax: +49 0 331 568
bres and the hypothetically separated Cellulose I crystals
3707. might not match the much more compliant matrix
E-mail address: ganster@iap.fraunhofer.de (J. Ganster). structure.

1359-835X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compositesa.2005.09.005
 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804 1797

Circumventing these problems while retaining the by an optical method. Conclusions will be drawn concern-
advantages mentioned above, it seems reasonable to ex- ing the potential applicability of high-tenacity cellulose
plore the reinforcing potential of man-made Cellulose fi- filament yarn as a reinforcing fibre for polypropylene and
bres spun in an industrial process like viscose or lyocell. other thermoplastic polymers.
First promising results from our laboratory, especially
with respect to strength and impact strength for rayon 2. Experimental
(viscose process) reinforced polypropylene were published
in [16]. Stiffness properties of compression moulded short 2.1. Materials
cut rayon reinforced polypropylene films as a function of
temperature were first investigated by Amash and 2.1.1. Cellulose filament yarn
Zugenmaier [17,18]. Both for isotropic and drawn films The cellulose high-tenacity filament yarn used in this
the authors conclude [18] that the influence of cellulose study was the rayon tyre cord yarn Cordenka 700 (Cord-
spun fibres on the viscoelastic behaviour of the films is enka GmbH, Germany) which is produced by a specially
greater then that of wood cellulose microfibres, which designed viscose process. Some important characteristics
were also investigated in this study. In the field of injec- are given in Table 1 and were measured on single filaments
tion moulding compounds, which is the topic of the pres- at 23 °C and 50% relative humidity with an initial clamp
ent paper, only one attempt to use a spun cellulosic fibre separation of 20 mm and a velocity of 10 mm/min. From
as polypropylene reinforcement has been found in the lit- the fibre titre of 1.8 dtex and the cellulose density of
erature (Paunikallio et al. [19,20]). The authors use vis- 1.5 g/cm3 the fibre diameter is 12 lm.
cose staple fibres (Säteri OY) of unspecified titre and
tenacity to reinforce a polypropylene homopolymer 2.1.2. Matrix polymers
(HD120MO, Borealis, MFI 8 g/10 min) and found tensile The matrix polymers used as well as some important
moduli and tensile strengths of up to 3.7 GPa and characteristics as given by the producers are listed in
70 MPa, respectively, for a rather high fibre load of Table 2. Other polypropylenes used merely for comparison
40 wt%. Best results were obtained using 6 wt% of maleic will be briefly specified in the text.
anhydride grafted polypropylene (Exxelor PO 1015) as a
coupling agent. Unfortunately, no results for impact 2.1.3. Coupling agents
strength are given. For coupling the cellulose fibres to the polypropylene
The purpose of the present paper is to explore the rein- matrix maleic anhydride grafted polypropylene (MAPP)
forcing capabilities of a high-tenacity rayon filament yarn has been used [18] with a high MFI (450 g/10 min at
(Cordenka 700) in thermoplastic matrices (predominantly 190 °C and 21.6 N) and a graft level >1 wt% the trade
polypropylene) for injection moulding applications. A name being Fusabond MD353D (Du Pont). MFI and
two-step processing technology has been developed [21] graft level for other MAPPs used for comparison age
to cut, incorporate and disperse the cellulose continuous given in Table 3. PE has been coupled with Fusabond
filaments in the thermoplastic matrix. Results will be E MB-100D (Du Pont) with and MFI of 2 g/10 min at
presented for mechanical properties, including impact 190 °C and 21.6 N and a graft level >1 wt%. For HIPS
strength, of injection moulded ISO test bars as a function Poly(styrene-co-maleic anhydride) with 7 wt% maleic
of fibre load, fibre cut length, matrix variations and varia- anhydride and MW 224000 (Aldrich) was used and
tions in coupling agent type and amount. Structural PLA has not been coupled at all. An amount of 3 wt%
features are being revealed by light microscopy and SEM, coupling agent with respect to the matrix polymer has
as well as the fibre length distribution in the final test bar been used in all cases unless otherwise stated in the text.

Table 1
Characteristics of Cordenka 700
Filament titre (dtex) Number of filaments per tow Modulus (cN/tex) (GPa) Tenacity (cN/tex) MPa Elongation at break (%)
1.8 1350 1300 19.5 55 825 13

Table 2
Matrix polymers used in this study and selected manufacturerÕs data
Trade name Abbreviation Producer MFI (g/10 min) Strength (MPa) Modulus (GPa)
Stamylan P 412MN40 PP Sabic 37 at 230 °C, 22 N 26 (yield) 1.55 (bend)
Hostalen GC 7260 PE Basell 23 at 190 °C, 5 N 30 (yield) 1.35 (tesile)
Lacqrene 4240 HIPS Atofina 4 at 200 °C, 50 N 26 (yield) 2 (bend)
PLA 2002D nature works PLA Cargill Dow 4–8 at 190 °C, 22 N 53 3.5
1798 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804

Table 3
DIN EN ISO 527 and 178, respectively, with a universal
Properties of MAPPs and PP used for testing the coupling efficiency (cf.
Fig. 3) testing machine (Zwick Z020). However, the tensile modu-
lus was determined as the maximum derivative at the
Abbreviation Trade name MFI Graft level
(g/10 min) (wt%) beginning of the stress–strain curve measured at 50 mm/
min testing speed except where indicated otherwise. In
A Fusabond MD353D 450a >1
(Du Pont) those cases, the modulus was measured according to ISO
B Fusabond M613-05 120a 0.4–0.7 527. Compression moduli were determined along the test
(Du Pont) bar axis with sample heights of 30 mm and a test speed
C Exxelor PO 1015 150a 0.42 of 0.3 mm/min. Double notched shear strength (DNS)
(Exxon)
experiments have been performed in compression with
D Exxelor PO1020 430a 0.68
(Exxon) 1 mm/min testing speed on an ISO test bar cut to 16 mm
PP used Stamylan P 83MF10 1.8b – length and notched (2 mm width) from either side to the
a
At 190 °C, 21.6 N. middle of the sample leaving a bridge of 2 mm to be
b
At 230 °C, 21.6 N. sheared. Charpy impact strengths of the composites were
determined with an impact tester (PSW 4J) according to
2.2. Methods DIN EN ISO 179 standard in the flatwise, unnoched, or
the edgewise notched (notch type A) modes. The test sam-
2.2.1. Compounding ples were conditioned at 23 °C and 50% relative humidity
A pultrusion technique was applied with a conventional for several days before testing and all the tests were per-
co-rotating twin screw extruder (Haake Rheocord 9000 formed under the same conditions.
PTW 25) equipped with a coating die assembly to cover
a number of (continuous) filament tows with the molten 2.2.4. Scanning electron microscopy
matrix-coupling agent mixture which was premixed before Cryo-fractured surfaces were generated by breaking the
fed into the extruder. PLA was thoroughly pre-dried in a test bars under liquid nitrogen conditions and subsequent
vacuum oven at 60 °C in the presence of P2O5 to avoid sputtering with Pt with a thickness of 4 nm. The fracture
water initiated degradation during extrusion. The maxi- surfaces were studied with an SEM Jeol JSM 6330 at 5 kV.
mum temperature of the extruder and the die were 200
and 195 °C, respectively. The coated yarns were cooled 2.2.5. Light microscopy
with water and cut into pellets, the length of which could The fibre distribution in the injection moulded speci-
be chosen between 3 and 5 mm. Then the pellets were dried mens was observed with a Jenalab polarization microscope
overnight at 110 °C for PP and PE, 80 °C for HIPS and in (Karl Zeiss, Germany) using 10 lm thick cuts prepared
vacuum at 60 °C in the presence of P2O5 for PLA. In a sec- perpendicular to the sample axis with a rotational micro-
ond step, the pellets were extruded with the same extruder tome HM 355 S (Microm International, Germany).
under the same conditions to homogenise the fibre-matrix
mixture. The screw configuration was chosen such that 2.2.6. Fibre length distribution
appropriate mixing elements were included to guarantee Cellulose fibres were isolated from the polypropylene
the dispersion of the fibres in the matrix. After cooling matrix using a Soxhlet extraction procedure with decalin.
the thread was cut into final pellets of defined lengths be- An extraction time of 16 h was necessary to remove all
tween 3 and 5 mm with diameters between 2 and 4 mm. the polypropylene, as verified by reaching a constant mass
Apparent pellet densities are above 300 g/l in all cases. for the extracted fibres. Subsequently, the fibres were rinsed
in decalin and vacuum dried for 24 h. The fibres were
2.2.2. Injection moulding spread on a microscope slide and their length was mea-
Standard test specimens were prepared according to DIN sured using the image processing system ‘‘Analysis’’ (SIS
EN ISO 527-2 (for tensile test) and DIN EN ISO 179 (for Ltd., Germany) with a ten fold magnification of a Jenalab
bending and Charpy impact test) using an injection mould- polarization microscope (Karl Zeiss, Germany). In that
ing machine (Allrounder 270 M 500-90, Arburg, Germany) way, approximately 1000 fibres were measured for each
with a ram pressure of 30 kN and a feed rate of 50 cm3/s. sample.
Zone temperatures were set to 170, 180, 190 and 200 °C from
feed to nozzle and the nozzle temperature was 200 °C, while 3. Results and discussion
the injection pressure was in the range from 400 to 700 bar.
The tool was designed for moulding two small standard test 3.1. Polypropylene
bars (Charpy and bending) and one dog bone shaped bar
(tensile) at a time, each with an edge gate. 3.1.1. Effect of coupling agent
Tensile strength and the double notched shear strength
2.2.3. Mechanical testing for Cordenka-PP composites with 20% fibre load as a
Tensile strength and modulus, and bending strength and function of standard coupling agent (MAPP) concentra-
modulus of the composites were measured according to tion are shown in Fig. 1. The effectiveness of the coupling
 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804 1799

25% fibre load shown in Fig. 2. On the left hand side no

MAPP is used and the fibres are pulled out with practically
no adherence to the matrix. On the right, by the use of 3%
MAPP, the fibres are covered with PP on the surface and
can transfer their reinforcing potential to the composite,
as it is also evident from Fig. 1. The small threads visible
in the picture consist of drawn matrix material.
A change in the molecular architecture of the MAPP (cf.
Table 3) has only a moderate influence on the properties of
the composites as shown in Fig. 3 for the strength of ex-
truded rods (no second cutting step and injection mould-
ing) as an example. Best results are obtained for the light
flowing, highly grafted product A, whereas both lower
MFI (higher molecular mass) and lower degree of grafting
Fig. 1. Tensile strength and double notched shear strength for Cordenka- for types B, C, and D give slightly lower values in accord
PP composites with 20% fibre load as a function of coupling agent with ref. [22].
(MAPP) concentration.
3.1.2. Fibre and fibre length distribution
is obvious: the tensile strength, e.g., is doubled by adding The high degree of homogeneity of the fibre distribution
the rather small amount of 1 wt% MAPP. After a plateau in the composite (25% fibre load) can be estimated from
between 1% and 3%, the properties begin to drop with
increasing concentration of MAPP. This is very likely to
be caused by the increasing amount of short PP chains in
the composite coming from the high MFI coupling agent.
In the following, 3 wt% of MAPP are chosen as a standard
and used throughout this work unless otherwise stated to
account for higher fibre loads which need higher amounts
of coupling agents [22].
The coupling mechanism between MAPP and cellulose
has been explained by Felix and Gatenholm [4] by the for-
mation of ester linkages between maleic acid anhydride and
the hydroxyl groups of the cellulose chains and by hydro-
gen bonds. Since in the present case the processing temper-
ature is above 170 °C, esterification should be the main
coupling mechanism. On the matrix side, simple physical
entanglement of the long PP segments of the MAPP in
the matrix PP provides for the anchorage of the ester
groups in the PP matrix.
The efficiency of the coupling is also demonstrated by Fig. 3. Tensile strength of extruded Cordenka-PP rods as a function of
the SEM photographs of cryo fractures of composites with MAPP type (cf. Table 3) with 20% fibre load.

Fig. 2. SEM cryo fracture photographs of Cordenka-PP composites with 25% fibre load and no coupling agent (left) and 3% MAPP (right).
1800 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804

Fig. 4. SEM picture of cut surface (left) and light microscopy picture of slice (right) cut perpendicular to test bar axis of Cordenka-PP composite (25%).

Fig. 4. On the left, an SEM picture of a surface of a testing tooth size 3 mm) do not grip the fibres and do not reduce
bar cut perpendicular to the melt flow direction and on the the length significantly, at least for the majority of the fi-
right a light microscopy picture of a 10-lm thick slice of the bres. On the other hand, with 5 mm starting pellet size,
same bar cut in the same direction are shown. There is an the cutting action of the extruder elements is clearly seen
even distribution of the fibres up to the surface of the bar by the maximum at around 2 mm. A second maximum is
(upper nearly horizontal line in the right picture) and only found between 4 and 4.5 mm. Such a double maximum
very few fibres stick out of the surface. This effect has not shape is expected for the fibre length distribution if an
been observed macroscopically, and might be due to the individual fibre is cut exactly once or not at all. In that
cutting process and therefore is an artefact. That means, case a symmetry axis at 2.5 mm should be present. This
a thin surface layer only consists of PP and thus makes is not the case indicating a multiple cut for a certain
the composite water resistant. amount of individual fibres.
In order to clarify the changes in fibre length during After pelletising the extruded rods with a cut length
the extrusion of the compound, homogenisation runs were identical to the initial pellet size from the first extrusion
performed with pellets of 3 and 5 mm cut length from the run, and injection moulding the compound into test speci-
first processing step (coating) giving extruded rods which mens, the fibre length distribution is changed into a shape
were not pelletised but analysed with regard to their fibre as shown in Fig. 6 for 3 mm (above) and 5 mm (below) cut
length distribution. The results are shown in Fig. 5 for the length, respectively. A considerable length reduction is ob-
3 mm (above) and the 5 mm (below) initial pellet size. served as compared to Fig. 5. There are few long fibres left
The width of the 3 mm distribution is likely to be due and no trace of a double maximum is detected. Fore the
to the tolerances in cut length of the starting pellets 3 mm case, the maximum of the distribution lies between
caused by a somewhat uneven action of the granulator. 250 and 500 lm, while for 5 mm cut length the maximum
Obviously, the mixing elements of the extruder (transverse is between 500 lm and 1 mm.

0.08
0.08
a) 3 mm initial length a) 3 mm cut
F r e q u en c y o f o c c u r e n c e

0.06 0.06
Frequency of occurence

0.04
0.03
0.02
0.00 0.00
0.06
0.020 b) 5 mm initial length
0.04 b) 5 mm cut
0.015
0.010 0.02
0.005
0.000 0.00
0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000
Fibre length [µm] Fibre length [µm]

Fig. 5. Fibre length distribution of extruded composite rods (25% fibre Fig. 6. Fibre length distribution of Cordenka-PP injection moulded bar
load) with 3 mm (above) and 5 mm (below) cut length in the first with 25% fibre load and 3 mm (above) and 5 mm (below) cut length in
processing step. both extruder runs.
 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804 1801

The mechanical properties, on the other hand, did not Table 4

change appreciably: strength and modulus increased by Fibre volume fractions (Vf), Reuss (EReuss), Voigt (EVoigt) and measured
tensile moduli (Eexp), fibre length efficiency factors gCox and orientation
approximately 5% for the 5 mm cut length vs. the 3 mm efficiency factors (go) for injection moulded Cordenka-PP test bars
and no difference was detected for Charpy notched and
Vf EReuss (GPa) Eexp (GPa) EVoigt (GPa) gCox go
unnotched impact strength.
0.16 1.63 3.23 4.29 0.942 0.70
0.2 1.72 3.57 5.10 0.946 0.65
3.1.3. Stiffness 0.25 1.81 3.80 5.92 0.950. 0.60
Tensile, bending, and compression moduli of Cordenka- 0.3 1.93 4.07 6.82 0.953 0.55
PP injection moulded test bars as a function of fibre load
are shown in Fig. 7. Here the tensile moduli were deter-
mined according to ISO 527. As expected, the moduli in- tween the lower and upper bounds, i.e., a mixed parallel–
crease with increasing fibre content. Compression moduli series or series–parallel model (Takayanagy models [23])
are considerably lower than the tensile counterparts, high- would be appropriate. However, in doing so the problem
lighting the anisotropic properties of the ‘‘soft’’ cellulose fi- to relate the second model fit parameter to a real structural
bres as compared to glass fibres. The latter act more like parameter must be solved.
stiffening rods, while the former can be pictured as rein- Since the finite fibres are embedded in the matrix, their
forcing ‘‘ropes’’. The bending moduli, in between the two effective moduli are reduced by a factor
others as expected, are rather close to the compression val-
gCox;L ¼ 1  tanhðbL=2Þ=ðbL=2Þ
ues, indicating a shift of the neutral plane separating com-
pression and tension regions in the sample towards the for a fibre of length L and radius R ([24,25]) and with
lower sample surface thus diminishing the thickness of 1=2 1=2
the layer loaded in tension. b ¼ ½2  Gm =ðEf  R2  lnðV f ÞÞ ;
For a composite material of two components, in the where Gm is the torsion modulus of the matrix and can be
present case fibre and matrix with moduli Ef and Em, set to Em/3 here. Using the fibre length distribution w(L),
respectively, there exist a lower and an upper bound to the final fibre length efficiency factor gCox is calculated by
the composite modulus Eexp as a function of the fibre Vol- Z
tan hðbL=2Þ
ume fraction Vf. The upper bound EVoigt corresponds to a gCox ¼ 1  wðLÞ dL.
parallel arrangement (constant strain) of the components bL=2
and is given by a simple rule of mixtures The Cox factors calculated for the fibre length distribution
EVoigt ¼ V f  Ef þ ð1  V f Þ  Em ; shown in Fig. 6(b) are given in Table 4. For the present set
of parameters, all the factors are fairly close to unity, i.e.,
whereas the lower limit (series coupling, constant stress) there are only minor reductions due to the finite length of
EReuss follows from a rule of mixtures of the reciprocals, the reinforcing fibres.
1=EReuss ¼ V f =Ef þ ð1  V f Þ=Em . A modified rule of mixtures taking into account both fi-
bre length distribution and fibre orientation through effi-
Both moduli as a function of fibre volume fraction are gi-
ciency factors gCox and go, respectively, can be set up as
ven in Table 4, using the measured matrix tensile modulus
follows [25]
of 1.39 GPa. Obviously, the measured values are in be-
Eexp ¼ gCox  go  V f  Ef þ ð1  V f Þ  Em .
Using the measured Eexp, and the calculated gCox, the ori-
entation efficiency factor go can be calculated as given in
Table 4. In contrast to glass fibres (Ref. [15], Fig. 10), the
orientation factor in the present case is a function of the fi-
bre load with reduced values for higher fibre concentra-
tions, indicating a mutual influence of the fibres
hindering the orientation formation. For fibre loads below
20% the values are comparable to the short glass fibre val-
ues of 0.69 ([15], Table 1) while for a concentration of 30%
the value drops to 0.55 being still well above the random
in-plane value of 0.38 [15].

3.1.4. Tensile and bending strength

The tensile strength and bending stress at 3.5% strain of
the outer fibre of Cordenka-PP injection moulded test bars
as a function of fibre load are shown in Fig. 8. The bending
Fig. 7. Tensile, bending, and compression moduli of Cordenka-PP stress at 3.5% strain of the outer fibre has been chosen as
injection moulded composites as a function of fibre load. the characteristic value for the bending behaviour since
1802 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804

90
out, a high level of 80–90 kJ/m2 is obtained, even at the
tensile lower temperature. The neat polypropylene does not break
80 bending at room temperature, while at 18 °C the impact strength
is (77 ± 13) kJ/m2. There is no obvious dependence on fibre
70
concentration similar to unnoched Izod values for glass fi-
Strength [MPa]

60 bre reinforced polypropylene ([15], Fig. 8). The glass fibre

reinforced sample from the present study (Piolen
50
G30CA60) gives 35 and 39 kJ/m2 for 23 and 18 °C,
40 respectively.
A clear trend to increased values for increased fibre con-
30
tent, similar to the behaviour with glass ([15], Fig. 7) is
20 found for the notched Charpy impact strength as demon-
strated in Fig. 10. Already for 24 wt% fibre load the values
0 10 20 30 40 50 are doubled at 23 °C and more than tripled at 18 °C as
Fibre Weight Fraction [%] compared to the neat resin. The rather poor impact behav-
Fig. 8. Tensile strength and bending stress at 3.5% strain of the outer fibre iour of the neat polypropylene, especially below its glass
of Cordenka-PP injection moulded composites as a function of fibre load. transition temperature of roughly 10 °C, is therefore dras-
tically improved by the addition of the cellulose man-made
fibres.
no break was observed in the experiments and the approx-
imations inherent in the calculations (simple beam theory)
lose their validity at higher deformations. For the rein- 3.1.6. Variation of polypropylene type
forced samples, the bending stress is generally lower then The influence of polypropylene chain architecture on
the tensile strength due to the better performance of the composite performance has been tested for 20% fibre load
material in tension then in compression (as also indicated and 3 wt% of standard coupling agent. As an example, re-
by the moduli in Fig. 7). For glass fibre polypropylene (Pio- sults for tensile and notched impact strength are given in
len G30CA60, Pio Co. Freiburg, Germany) this behaviour Fig. 11 for the standard block copolymer (Copo, cf. Table
is inverted, the tensile strength is 59 MPa and is thus lower 2), a random copolymer (RA-Copo, Stamylan P
than the bending stress, which is 71 MPa. The same re- RA12MN40, DSM Co.), a homopolymer (Homo, Stamy-
versed order is seen for the unreinforced sample. In both lan P 112MN40, DSM Co.), and a PP homopolymer pow-
latter cases the stiffness to strength ratio is higher and thus der (Eltex P HV001PF, Solvay Co.). The overall picture
the bending experiments give higher stress values than for does not change very much: both properties are improved
the cellulose composite case. considerably. A slight influence of the matrix properties
seems to be transferred to the composite. For a specially
3.1.5. Impact behaviour impact modified PP (Stamylan P 83MF10) with a low melt
The Charpy impact strengths (flatwise, unnotched) of flow index (1.8 g/10 min at 230 °C and 21.6 N) strength
Cordenka-PP injection moulded test bars as a function of was tripled, but notched impact strength could not be im-
fibre load at 23 and 18 °C are shown in Fig. 9. Through- proved, similar to the picture with HIPS (see below).

100
14
Room temperature
Notched Charpy [kJ/m ]

-18˚C
2

12
90
Charpy [kJ/ m ]
2

10

80 8

70 Room temperature 4
-18˚C
2

60 0
20 25 30 35 40 45 0 10 20 30 40 50
Fibre Weight Fraction [%] Fibre Weight Fraction [%]
Fig. 9. Charpy impact strength (flatwise, unnotched) of Cordenka-PP Fig. 10. Charpy impact strength (edgewise, V-notch) of Cordenka-PP
injection moulded composites as a function of fibre load at 23 and 18 °C. injection moulded composites as a function of fibre load at 23 and 18 °C.
 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804 1803

70
3.2. Alternative matrices
20
60 For alternative matrices to polypropylene exploratory
50 experiments have been performed in order to test the rein-

Charpy (n) [kJ/m ]

2
15
Strength [MPa]

forcement capabilities of the cellulose man-made fibre. Re-

40 sults for selected mechanical properties of the composites
30 10 using the coupling agents described in Section 2.1 in com-
parison to the neat resins are presented in Table 6. Except
20 for the PLA case, the same coupling scheme, i.e., ester link-
5
10 ages through MAH and cellulose hydroxyl groups on the
fibre side and physical entanglements on the matrix side,
0 0
Copo RA-Copo Homo Powder has been employed.
Polypropylene type
For PE the system works remarkably well. Tensile mod-
ulus is almost tripled, strength and notched impact strength
Fig. 11. Tensile strength and notched Charpy impact strength for various are more than tripled and increased 5-fold, respectively.
neat resins (open symbols) and Cordenka-PP injection moulded compos-
The HIPS composite shows a modulus increase of 50%
ites with 20% fibre load (full symbols).
and a doubled tensile strength when compared to the pure
HIPS. The unnotched and notched Charpy impact
3.1.7. Comparison with short glass fibre reinforced strengths are reduced to 75% and 88%, respectively. It
polypropylene seems that the reinforcement with the cellulose fibres inter-
In order to estimate the performance of the composites feres with the impact modification inherent in the polysty-
developed here, a comparison of some mechanical proper- rene. However, trials with unmodified polystyrene were not
ties of the composite with 30 wt% fibre to those of a 30 wt% successful, since with the current processing method the
short glass fibre reinforced PP (Piolen G30CA60) is given coated fibre bundle could not be cut due to the very brittle
in Table 5. While strength and unnotched impact strength nature of the polymer. Moreover, the coupling agent might
are superior for the cellulose composite, notched impact not be optimal in two respects. First, with 7 wt% maleic
strengths at room temperature and lowered temperature acid in the chain, the remaining polystyrene segments
are similar, and the modulus is lower by approximately might be too short to entangle properly in the polystyrene
25%. This behaviour can be explained by the anisotropic matrix. Second, due to the copolymer nature the accessibil-
nature of the cellulose fibre which acts like a reinforcing ity of the maleic anhydride groups could be less than for
rope in contrast to the reinforcing bars for the glass fibre grafted entities.
case. Studies not reported here, however, show that the A third alternative matrix material investigated is the
stiffness can be improved significantly by adding lignocellu- biogenic and biodegradable poly(lactic acid) (PLA) allow-
losic or Lyocell fibres or inorganic fillers, without seriously ing the composite to be fully biodegradable. Modulus
compromising the remaining properties, notably the high and strength increase by 50%. In this system, the fibre
impact strength. works as a strong impact modifier: Charpy unnotched

Table 5
Comparison of selected mechanical characteristics of glass fibre reinforced PP (GF-PP, Piolen G30CA60) and Cordenka-PP composite with 30 wt% fibres
(PPRayCo30)
Property Unit Value for GF-PP Value for PPRayCo30
Tensile strength MPa 56.3 ± 0.4 78.7 ± 1.1
Tensile modulus GPa 4.1 ± 0.2 2.94 ± 0.04
Charpy unnotched impact strength, 23 °C kJ/m2 35 ± 3 89 ± 8
Charpy unnotched impact strength, 18 °C kJ/m2 39 ± 3 88 ± 6
Charpy notched impact strength, 23 °C kJ/m2 13.4 ± 0.3 12.1 ± 0.7
Charpy notched impact strength, 18 °C kJ/m2 8.5 ± 0.3 9.3 ± 1.5

Table 6
Selected mechanical properties of injection moulded test bars of neat resins and composites with 25% Cordenka fibres
Material Modulus (GPa) Strength (MPa) Charpy(u) (kJ/m2) Charpy(n) (kJ/m2)
PE 1.10 ± 0.03 23.6 ± 0.5 n.b. 3.9 ± 0.1
PE composite 3.0 ± 0.2 82 ± 1 79 ± 6 22 ± 1
HIPS 2.0 ± 0.1 28.9 ± 0.2 58 ± 15 12.3 ± 0.5
HIPS composite 3.1 ± 0.4 58 ± 3 45 ± 6 10.8 ± 0.4
PLA 2.80 ± 0.04 71.0 ± 0.6 14.4 ± 1.6 2.8 ± 0.1
PLA composite 4.2 ± 0.1 108 ± 2 69 ± 8 8.4 ± 0.4
1804 J. Ganster et al. / Composites: Part A 37 (2006) 1796–1804

and notched impact strengths are increased by 380% and [6] Wambua P, Ivens J, Verpoest I. Natural fibres: can they replace glass
200%, respectively. SEM pictures reveal that the interfacial in fibre reinforced plastics?. Comp Sci Technol 2003;63:1259–64.
[7] Zafeiropoulos NE, Baillie CA, Matthews FL. A study of transcrys-
adhesion between fibre and PLA matrix is moderate to tallinity and its effect on the interface in flax fibre reinforced
poor. This was to be expected since no coupling agent composite materials. Comp A 2001;32:525–43.
has been used. In contrast to the PP, PE and HIPS cases, [8] Karmaker AC, Youngquist JA. Injection moulding of polypropyl-
no commercial coupling agents for PLA, e.g., an MAH ene reinforced with short jute fibres. J Appl Polym Sci
grafted PLA, are available. 1996;62:1147–51.
[9] Rana AK, Mandal A, Mitra BC, Jacobson R, Rowell R, Banerjee
AN. Short jute fibre-reinforced polypropylene composites: effect of
4. Conclusions compatibilizer. J Appl Polym Sci 1998;69:329–38.
[10] Khan MA, Hinrichsen G, Drzal LT. Influence of novel coupling
High tenacity cellulose man-made filament yarn usually agents on mechanical properties of jute reinforced polypropylene
employed for tyre reinforcement has proven to be an composite. J Mater Sci 2001;20:1711–3.
[11] Joseph PV, Joseph K, Thomas S. Effect of processing variables on the
appropriate reinforcing material for commodity thermo- mechanical properties of sisal-fibre-reinforced polypropylene com-
plastics such as polypropylene. Highly homogeneous injec- posites. Comp Sci Technol 1999;59:1625–40.
tion moulding compounds can be produced by a two step [12] Chu PF. Glass-fiber reinforced polypropylene. In: Karian HG, editor.
pultrusion method. The overall mechanical property level Handbook of polypropylene and polypropylene composites. New
is comparable with short glass fibre reinforced polypropyl- York: Marcel Dekker; 2003. p. 311–81.
[13] Specht K, Bledzki AK, Thole V, von Buttlar HB. Potential of refined
ene (GF-PP) with advantages in strength and impact hemp in injection moulded polypropylene composites. In: Proceed-
strength. The stiffness of the composites is lower than for ings of the 5th global wood and natural fibre composites symposium,
GF-PP due to the anisotropic nature of the fibre acting Kassel, April; 2004. p. A8-1–18.
as a reinforcing rope rather than a reinforcing bar. Com- [14] Khan MA, Ganster J, Fink H-P. Natural and man-made cellulose
pared to glass fibres, the cellulose yarn has a lower density fibre reinforced hybrid polypropylene composites. In: Proceedings of
the 5th global wood and natural fibre composites symposium, Kassel,
(1.5 g/cm3 vs. 2.5 g/cm3), is not abrasive to the processing April; 2004. p. A19-1–8.
equipment and can be combusted without residue. Other [15] Thomason JL. The influence of fibre length and concentration on the
thermoplastics like polyethylene, high impact polystyrene properties of glass fibre reinforced polypropylene: 5. Injection
and poly (lactic acid) (PLA) have been successfully rein- moulded long and short fibre PP. Comp A 2002;33:1641–52.
forced as well. Especially the brittleness of PLA is drasti- [16] Weigel P, Ganster J, Fink H-P, Gassan J, Uihlein K. Polypropylene-
cellulose compounds – high strength cellulose fibres strengthen
cally reduced by the incorporation of the cellulose injection moulded parts. Kunststoffe plast europe 2002;92:35–7.
filaments leading to a completely biogenic and biodegrad- [17] Amash A, Zugenmaier P. Study on cellulose and xylan filled
able composite with excellent mechanical properties. polypropylene composites. Polym Bull 1998;40:251–8.
[18] Amash A, Zugenmaier P. Morphology and properties of isotropic
Acknowledgement and oriented samples of cellulose fibre-polypropylene composites.
Polymer 2000;41:1589–96.
[19] Paunikallio T, Kasanen J, Suvanto M, Pakkanen TT. Influence of
This work is part of the internal Fraunhofer project maleated polypropylene on mechanical properties of composite made
‘‘Novel polymer-based commodity materials’’. of viscose fibre and polypropylene. J Appl Polym Sci 2003;
87:1895–900.
References [20] Paunikallio T, Suvanto M, Pakkanen TT. Composition, tensile
properties, and dispersion of polypropylene composites reinforced
[1] Bledzki AK, Gassan J. Composites reinforced with cellulose based with viscose fibres. J Appl Polym Sci 2004;91:2676–84.
fibres. Prog Polym Sci 1999;24:211–74. [21] Gassan J, Einsiedel R, Fink H-P, Weigel P. Profiled part and
[2] Gauthier R, Joly C, Coupas AC, Gauthier H, Escoubes M. Interfaces aggregates for making same. WO 03/033227.
in polyolefin/cellulosic fiber composites: chemical coupling, morphol- [22] Snijder MHB, Bos HL. Reinforcement of poylpropylene by annual
ogy, correlation with adhesion and aging in moisture. Polym Comp plant fibres: optimisation of the coupling agent efficientcy. Comp
1998;19:287–300. Interfaces 2000;7:69–75.
[3] Coutinho FMB, Costa THS. Performance of polypropylene-wood [23] Ward IM, Hadley DW. An introduction to the mechanical properties
fibre composites. Polym Testing 1999;18:581–7. of solid polymers. Chichester: Wiley; 1993, p. 154.
[4] Felix JM, Gatenholm P. The nature of adhesion in composites of [24] Cox HL. The elasticity and strength of paper and other fibrous
modified cellulose fibres and polypropylene. J Appl Polym Sci materials. Br J Appl Phys 1952;3:72–9.
1991;42:609–20. [25] Bader MG, Hill AR. Short fiber composites. In: Chou T-W, editor.
[5] Felix JM, Gatenholm P, Schreiber HP. Controlled interactions in Material science and technology – a comprehensive treatment.
cellulose-polymer composites. I: effect on mechanical properties. Structure and properties of composites, vol. 13. Weinheim: VCH;
Polym Comp 1993;14:449–57. 1993. p. 291–338.