Composites: Part B 60 (2014) 603–611

Contents lists available at ScienceDirect

Composites: Part B

journal homepage: www.elsevier .com/locate /composi tesb

Development of polyhydroxyalkanoates/poly(lactic acid) compositesreinforced with cellulosic fibers

1359-8368/$ - see front matter � 2014 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.compositesb.2014.01.001

⇑ Corresponding author. Tel.: +351 962872974.E-mail addresses: nuno.loureiro@fe.up.pt (N.C. Loureiro), jesteves@fe.up.pt

(J.L. Esteves), jcv@dep.uminho.pt (J.C. Viana), sghosh@dep.uminho.pt (S. Ghosh).

N.C. Loureiro a,⇑, J.L. Esteves a, J.C. Viana b, S. Ghosh b

a University of Porto, Faculty of Engineering, Mechanical Engineering Department, R. Dr. Roberto Frias, s/n, 4200-465 Porto, Portugalb IPC/I3N – Institute for Polymers and Composites, Polymer Engineering Department, University of Minho, Campus de Azurém, 4800-048 Guimarães, Portugal

a r t i c l e i n f o

Article history:Received 20 January 2013Received in revised form 26 November 2013Accepted 3 January 2014Available online 9 January 2014

Keywords:E. Thermoplastic resinB. Mechanical propertiesC. Analytical modellingD. Mechanical testingB. Thermal properties

a b s t r a c t

In this work, the mechanical behavior of [30:70] (wf) PHA/PLA blend matrix reinforced with cellulosicfibers is investigated in the range of fiber incorporation of 10%, 20% and 30% wf.

The mechanical properties of the composites can be optimized through the variation of the fiber con-tent on the composite. It is possible to predict the tensile properties recurring to modified Halpin–Tsaiequation, Ishai and Cohen Model and Rule-of-Mixtures Models.

Analyzing the data from the dart impact test, the irregular fiber distribution for fiber ratio over than20% is confirmed. The highest synergetic effect is found for fiber ratio of 20%.

The incorporation of fiber improves the Heat Deflection Temperature (HDT) of the matrix when thefiber presents a homogeneous distribution through the matrix.

The microscopic analysis allows identifying the good dispersion of the fiber through the matrix and alower interfacial adhesion.

The incorporation of 20% wf of fibers drives to an increasing of the tensile, flexural and impact behaviorand also to a superior HDT.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Fiber-reinforced polymers have successfully proven their valuein various applications due to their good properties. Commonly,petrol-based polymers are reinforced with glass or carbon fibersoffering advantageous mechanical properties at low weight. To re-place these petrol-based polymers and inorganic fiber reinforce-ments, natural fiber polymer composites technology is focusedon creating more sustainable high performance and lightweightmaterials [1–5]. In fact, the growing of the environmental aware-ness and the creation of the new standards including ‘‘End of LifeVehicle’’ (ELV) regulations in the EU automotive sector forces thestudy and production of more environmental friendly and biode-gradable materials. Composite materials based on raw materialsderived from natural renewable resources are being studied, anew class of eco-composites.

Polymers from renewable resources, suitable for environmentalfriendly and biodegradable composite matrices, such as soy-oilbased epoxy, starch based polymers, polycaprolactone (PCL), poly-hydroxy butyrate (PHB), polylactic acid (PLA), have been investi-gated in the unfilled state [6–11]. PLA is a polymer produced by

the fermentation of simple sugars such as glucose and maltosefrom corn or potato, sucrose from cane or beet sugar and lactosefrom cheese [12]. It is a linear aliphatic polyester thermoplastic,used as package materials and in production of cloths, carpet tiles,surgical and biomedical applications among others. The PLAmechanical properties have been reported in the literature. PHAis a generic designation of polyester polymers produced by thebacterial fermentation of sugars and lipids. These polyesters are acarbon storage and energy reserves in bacteria such as RalstoniaEutropha, Bacillus Megaterium, and Azotobacterchroococum, andhave a wide range of mechanical properties (e.g., strength andYoung’s Modulus). Besides these neat polymers, blends of poly-mers from renewable resources have been investigated enlargingtheir range of properties, e.g., PHA/PLA blends [12–14].

The reinforcement with cellulosic fibers gives to the polymercomposite relatively good mechanical properties (stiffness,strength, and toughness) as well as, a low cost substitute solutionsfor some composites reinforced with glass synthetic fibers, and anease of disposal. Several studies have been made using polymersreinforced with natural fibers such as, jute, flax, banana, sisal, pine-apple, coir and oil palm [15–17]. Cellulosic fibers made from woodhave also being used, mainly those coming from paper pulp indus-try [18]. Lignocellulosic fibers present some advantageous: lowdensity, high specific strength and modulus, nonabrasive character,acoustic insulation, and low cost. But also some disadvantages:

Table 1Raw-material manufacturer’s datasheet properties values.

PLA PHA

Melt temperature (�C) 188–210 145–155Degradation temperature (�C) 200 200Tensile strength at break (MPa) 48 35Tensile elongation at break (%) 2.5 2Tensile modulus (MPa) 2700 2950HDT A (1.8 MPa) (�C) 60 72.5Density (kg/m3) 1240 1250MFI (190 �C/2.16 kg) (g/600 s) 30–40 15–30

604 N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611

high moisture absorption, inherent quality variations, hydrophiliccharacter (and poor compatibility with the hydrophobic polymers),and odor development in application.

The behavior of any composite material depends directly on thefiber content, on the degree of dispersion and orientation of the fi-bers, on matrix and fiber properties and on the degree of interac-tion/adhesion between the matrix and reinforcing phase [20].Guimarães [19] stated that for PLA/cellulosic fibers composites isnot possible to incorporate more than 30% of fiber, otherwise thePLA matrix will not be able to accommodate all the fibers. Process-ing conditions play also an important role on the properties ofthese eco-composites. Investigations on the effect of processingtechniques on the properties of composites [21,22] concluded thattwin screw extrusion resulted in better fiber dispersion in a poly-meric matrix, but a pre-processing stage before any molding pro-cess (e.g., injection molding) causes material degradation andreduction on the composite properties.

In a previous work [23] we had investigated the optimization ofthe ratio PHA/PLA in the blend. A [PHA:PLA] ratio of [30:70] wt%showed the best mechanical performance. In this work we proceedwith our aim of improving the performance of eco-composites,envisaging automotive interior applications (e.g., door trims), byreinforcing the 30PHA:70PLA matrix with cellulosic fibers. Thisstudy presents the mechanical properties of these eco-compositeswith different percentage of incorporation of cellulosic fibers up to30% wf.

2. Mechanical properties prediction models

2.1. Modified Halpin–Tsai equation (mHT)

The tensile properties of the composites (indicated by subscriptc) can be predicted by the modified Halpin–Tsai equation, recur-ring to matrix (indicate by subscript m) and of fibers (indicate bysubscript f) properties. This model, as referenced by Sathyanaraya-na [14], is used in determining the properties of composites thatcontain discontinuous fibers. Latter, Nielson modified this equationincluding the maximum packaging fraction, Umax, of the reinforce-ment. The modified Halpin-Tsai equation, presented bySathyanarayana, gives the modulus, Ec, and the tensile stress, rc,of the composite, respectively, by [14]:

Ec ¼ Em1þ ngeVf

1� gewV f

� �ð1Þ

rc ¼ rm1þ ngtVf

1� gtwV f

� �ð2Þ

where w depends upon the particle packing fraction, n is deter-mined from the Einstein coefficient f, and /max is the maximumpacking fraction and has a value of 0.82 for random packing of fibers[14]:

w ¼ 1þ 1� /max

/2max

!ð3Þ

n ¼ f� 1 ð4Þ

f ¼ 1þ 2ld

ð5Þ

ge

EfEm� 1

EfEmþ n

ð6Þ

gt ¼rfrm� 1

rfrmþ n

ð7Þ

The Einstein coefficient f, is determinate recurring to the averagelength (l) and diameter (d) of the fibers.

2.2. Ishai and Cohen Model (ICm)

For an approximate solution, Ishai and Cohen [14] assumes thatthe constituents are in a state of macroscopically homogeneousstress. Adhesion is assumed to be maintained at the interface ofa cubic inclusion embedded in a cubic matrix. When a uniform dis-placement is applied at the boundary the elastic modulus of thecomposite is given by the following equation:

Ec ¼ Em 1þ V f

m=ðm� 1Þ � V1=3f

!ð8Þ

m ¼ Efib

Emð9Þ

2.3. Rules of mixtures

The rule of mixtures (ROM) considers a perfect adhesion be-tween the matrix (indicated by subscript m) and the perfect dis-persed reinforce fiber (indicated by subscript f) into the matrix.This model can be used to predict the initial modulus and the flex-ural stress of the composite by, respectively:

Ec ¼Ef

Em� 1

� �� V f þ 1

� �� Em ð10Þ

rc ¼rf

rm� 1

� �� V f þ 1

� �� rm ð11Þ

where Em is the initial modulus of the matrix, Ef the initial modulusof the fibers, Ec the modulus of the composite, Vf the volume frac-tion of the fiber, rf the maximum stress of the fiber, rm the maxi-mum stress of the matrix, and rc is the maximum stress of thecomposite.

3. Materials and methods

3.1. Materials

The polymers used in this work were:

� PHA, under the trade name PHI002, manufactured byNatureplast,

� PLA, under the trade name Ingeo Biopolymer 3251D, man-ufactured by NatureWorks LLC.

The material suppliers provided the data properties presentedin Table 1.

The cellulosic fibers are from Portuguese Eucalyptus globulustrees, and becomes from a Portuguese paper pulp plant. The

N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611 605

eucalyptus fibers are made of pure cellulose from the bleachedkraft pulping process, being the raw materials for high qualitypaper production. The white cellulosic fibers used in this workwere pickup at the end of the chemical treatment that anticipatesthe paper production. The main properties of these fibers arepresented in Table 2.

3.2. Preparation of the composites

The materials were dried at 60 �C for 24 h before processing andkept into separate Ziploc bags. The composites were compoundedin a twin screw Coperion Extruder (Werner & Pfleiderer), withdifferent weight percentages of fibers of 0%, 10%, 20% and 30%.The extruder temperature profile used is shown in Fig. 1. Thistemperature profile was obtained comparing the injection profileobtained with PLA/PHA blends and with other works involvingcellulosic fibers.

The different ratios (polymers and fibers) were adjusted by set-ting the feed throat of each hopper. The used mass flow rate of eachextruded composite is expressed into Table 3.

Where QPHA, QPLA, QFiber and Qtotal are the massic flow rate ofPHA, PLA, cellulosic fiber and total massic flow rate of the eco-com-posite, respectively. After the composite compounding, the mate-rial is pelletizer, and then the pellets are stored in Ziploc bagsuntil starting of the injection molding process. The composite pel-lets were then injection molded in a Ferromatik Milacron K85injection machine, being produced moldings of different geome-tries (see below). The mold temperature was set at 20 �C, and theinjection temperature profile is presented in Fig. 2.

This temperature profile was established by combining themelting temperature of the polymers, the degradation temperatureand the injection molding conditions suggested by the supplier’sdatasheets. A constant injection velocity of 20 mm/s (correspond-ing to an injection flow rate of 6.3 cm3/s) was used.

The several composite were injection molded in the form ofspecimens with the following geometries and dimensions (accord-ing to the respective standard):

� Flexural and HDT specimens: 12 � 150 � 6 (mm) parallel-epipedic bars,

� Impact specimens: central gated discs of Ø60 � 2 (mm),� Tensile specimens (type II): dog-bone geometry with a nar-

row section of 57 � 10 � 4 (mm), and an overall dimen-sions of 183 � 19 � 4 (mm).

3.3. Mechanical characterization

3.3.1. Tensile properties measurementsThe tensile properties were tested in a universal mechanical

testing machine Shamidzu AG-X 100 kN, equipped with a 50 mmShamidzu extensometer, according to ASTM D638 test procedures.The crosshead velocity used was of 5 mm/min, and the tests wereperformed in a standard laboratory atmosphere of 23 ± 2 �C and50 ± 5% relative humidity. A grip distance of 150 mm was used.The envisaged tensile properties were the initial modulus, the

Table 2Eucalyptus globulus properties [24,25].

Average fiber diameter (lm) 10.9Average fiber length (mm) 0.66Tensile strength at break (MPa) 160Tensile elongation at break (%) 5.2Tensile modulus (GPa) 17.4Flexural modulus (GPa) 16Flexural strength at break (MPa) 130Density (kg/m3) 1600

maximum/yield stress and the strain at break. At least 11 speci-mens were tested for each reinforced blend composition.

3.3.2. Flexural properties measurementsA Universal Tiratest 2705 5 kN Machine was used to measure

the flexural properties according to ASTM D790 standard. It hasbeen used a 3-point flexural test, with a crosshead velocity of2.56 mm/min and a spam of 96 mm. Tests were performed in astandard laboratory atmosphere of 23 ± 2 �C and 50 ± 5% relativehumidity. The envisaged flexural properties were the initial modu-lus, the maximum stress and the strain at maximum stress. At least11 specimens were tested for each reinforced blend composition.

3.3.3. Instrumented Impact properties measurementsInstrumented impact tests are performed according ISO 6603-2

standard in a CEAST Fractovis plus Impact machine (Impact Energyof 12,5J) on central gated discs of Ø60 � 2 (mm). All tests were car-ried out in standard laboratory atmosphere of 23 ± 2 �C and 50 ± 5%relative humidity. From the force–displacement curve, the impacttoughness was calculated as the area below the force–displace-ment graph. The impact data presented are the average of 7measurements.

3.3.4. Heat Deflection Temperature (HDT) measurementsThe Heat Deflection Temperature (HDT) was measured accord-

ing to ISO 75-2, in a RAY-RAN HDT apparatus. This test used themethod HDT A with an applied stress state of 1.8 MPa and anincreasing temperature rate of 120 �C/h. Presented HDT resultsare the average values of three measurements.

3.4. Microscopy studies

3.4.1. Optical microscopyThe fiber distribution in the composite was investigated by

optical microscopy in an Axiophot optical microscope from Zeissequipped with an AxioCamICc 3. The all surface of the sampleswere analyzed.

3.4.2. Scanning electron microscopyThe fiber distribution in the composite and the adhesion be-

tween the fiber and the matrix were investigated by SEM in anEDAX-Pegasus X4M electronic microscope. The samples werecoated with a thin layer of gold using a Quorum/Polaron E6700high vacuum evaporator.

4. Results and discussion

4.1. Tensile behavior

The results of the tensile tests of the composites are given inTable 4. An experimental stress–strain demonstrative curve foreach specimen is shown into Fig. 3. The horizontal step representsthe time of extensometer removal during the test.

Based on the predictive models, the estimated mechanical prop-erties are presented in Table 5.

Fig. 4 shows the evolution of the modulus, E, with the reinforcedfiber weight fraction.

The increase of fiber content results in a general increment onthe initial tensile modulus of the composite. This is expected be-cause the fiber contributes to the stiffness of the final composite.In Fig. 4 are also presented the predictions of E based on the abovementioned models: ROM, mHT and ICm.

ROM model supposes that the fiber and the matrix present aperfect adhesion. The ICm is based in a cubic approximation. Theresults show that this model provides the lowest predication,

Fig. 1. Compounder temperature profile.

Table 3Extrusion flow rates.

Qtotal

(kg/h)QPHA

(kg/h)QPLA

(kg/h)QFiber

(kg/h)

100% Matrix (PHA/PLA blend) 4 1.2 2.8 090% Matrix/10% fiber 4 1.08 2.52 0.480% Matrix/20% fiber 4 0.96 2.24 0.870% Matrix/30% fiber 1 0.21 0.49 0.3

Fig. 2. Injection molding temperature profile.

Table 4Tensile properties of PHA/PLA composites reinforced with cellulosic fiber.

Fiber ratio (massfraction) (%)

Tensilemodulus (GPa)

Maximumstress (MPa)

Strain at maximumstress (%)

0 3.36 ± 0.07 40.00 ± 1.5 2.0 ± 0.0610 4.29 ± 0.37 44.21 ± 1.4 2.7 ± 0.3320 5.14 ± 0.31 49.01 ± 1.6 2.4 ± 0.3130 10.01 ± 0.34 79.72 ± 1.2 1.9 ± 0.18

Table 5Predicted and experimental tensile modulus.

Reinforce fiber(mass fraction) (%)

Tensile modulus (GPa)

Experimental ROM mHT ICm

0 3.36 ± 0.07 3.36 3.36 3.3610 4.29 ± 0.37 4.72 4.34 3.6920 5.14 ± 0.31 6.09 5.38 4.1430 10.01 ± 0.34 7.45 6.47 4.72

Fig. 4. Experimental tensile modulus, E, results and predicted values from themicromechanical models.

Table 6Experimental and predicted tensile stresses.

Reinforce fiber(mass fraction) (%)

Tensile stress (MPa)

Experimental ROM mHT

0 40.00 ± 1.5 40.00 40.0010 44.21 ± 1.4 49.54 44.8320 49.01 ± 1.6 59.52 49.9130 79.72 ± 1.2 69.97 52.24

606 N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611

meaning that the based hypothesis of the model is not suitable forthis specific composite. mHT equation supposes that the fiber pre-sents a homogenous distribution through the matrix. The mHTequation gives an excellent prediction of E for incorporation until20% (wf) of cellulosic fibers. Analyzing only until 20% (wf) of fiberincorporation, the maximum deviation between mHT equation andthe experimental values is only about 3.5%, and for the ROM this

Fig. 3. Tensile stress–strain curves obtained experimentally.

Fig. 5. Tensile maximum stress results and predicted values.

Table 7Flexural properties of PHA/PLA composites reinforced with cellulosic fiber.

Fiber ratio (massfraction) (%)

Flexuralmodulus (GPa)

Maximumstress (MPa)

Strain at maximumstress (%)

0 3.09 ± 0.13 77.49 ± 1.54 4.00 ± 0.2910 4.44 ± 0.07 82.51 ± 1.22 3.14 ± 0.2220 5.59 ± 0.09 89.02 ± 1.36 2.78 ± 0.1230 6.35 ± 0.19 85.36 ± 1.57 2.11 ± 0.04

Fig. 7. Flexural modulus, Ef, results and predicted values from model.

Fig. 8. Flexural maximum stress results and predicted values.

N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611 607

deviation reaches 18%. This anticipates a homogenous dispersionof the fiber through the matrix. For 30 wt% of fibers, the modelsunderpredict the experimental value. This may result from severalfactors including, not only the data obtain from the fibers, but alsofrom the geometric assumptions from the adopted models.

The incorporation of more than 20% of fibers will lead to a non-homogeneous composite. For fiber incorporation of more than 20%,the amount of fibers, allied with the flow caused by the injectionprocess and with the sample geometry, will start to present a floworientation.

Since the presented models are valid for random fiber displace-ment the induced orientation leads to a divergence between theexperimental data (flow orientated) and the models (randomorientation).

In Table 6 are presented the experimental values and the pre-dictions of the maximum tensile stress for ROM and mHT modelsfor the various fiber incorporation ratios.

Fig. 5 shows the variations of the maximum tensile stress withthe wt% of fiber and respective models predictions.

Fig. 6. Flexural stress–strain curv

The maximum stress of the composites can be also estimatedwith a good agreement from the above presented prediction mod-els, namely the mHT equation. This equation gives very good pre-dictions for incorporation of fibers until 20 wt%. Until thispercentage, the maximum deviation between mHT equation andthe experimental values is about 1.3%, and for the ROM is about20%. This result anticipates a homogenous dispersion of the fiberthrough the matrix. Again, for 30 wt% of fibers, the models underpredict the experimental value.

es obtained experimentally.

Fig. 9. Experimental impact force versus time for the instrumented impact tests.

Table 8Composite impact properties on central gated discs of £60 � 2 (mm).

Fiber ratio (mass fraction) (%) Impact energy (J) Deflection at break (mm)

0 1.7 ± 0.2 4.6 ± 1.410 2.8 ± 0.5 4.8 ± 0.420 2.8 ± 0.2 4.4 ± 0.830 2.3 ± 0.3 3.2 ± 0.8

Fig. 10. Impact toughness.

Fig. 11. Impact maximum deflection of various eco-composites.

Table 9Heat-Deflection Temperature of composites.

Fiber ratio (mass fraction) (%) HDT (�C)

0 48.5 ± 0.910 49.2 ± 0.620 56.0 ± 0.230 51.7 ± 0.2

608 N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611

Fig. 12. Experimental HDT evolution over fiber composition.

Fig. 13. Optical microscopy analysis (m

Fig. 14. SEM analysis; magnification: 1000�; fracture surface; (TOP LEFT, only matrix; T

N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611 609

4.2. Flexural behavior

The flexural test results of the composites are given in Table 7.An experimental stress–strain demonstrative curve for each speci-men is depicted in Fig. 6.

Fig. 7 shows the variations of Ef with% of fiber and respectivevalues of the prediction model (ROM). As for the tensile modulus,the ROM suggests that for fiber incorporation superior to 20% thefiber distribution is non-homogeneous. Conversely, to the tensilemodulus, the ROM models gives acceptable predictions up to 30%of incorporation of fibers.

Fig. 8 shows the variations of maximum flexural stress with% fi-ber and respective models predictions. Again, the increment on thefiber content leads to a general increasing of the maximum flexuralstress of the composite. However the incorporation of more than20% of fiber results on a reduction on the maximum flexural stress,and a divergence from the theoretical predictive values. In general,

agnification 20�); Sample Surface;

OP RIGHT 10% wf Fiber; BOTTOM LEFT 20% wf fiber; BOTTOM RIGHT 30% wf fiber).

Fig. 15. Details of the interface matrix-fiber obtain in SEM analysis (20% wf fiber) magnification: 1500� (left) 10.000� (right); fracture surface.

610 N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611

the ROM gives very good predictions for fiber incorporation levelsbelow 20%.

4.3. Impact behavior

Fig. 9 presents the impact force over time during the impact testof the composites. The incorporation of fiber improves the energyabsorption capabilities of the composites. However due to thenon-homogeneous distribution of the fiber in the 30% fiber-com-posite, the impact toughness appears to decrease because the fiberbundles may act as stress concentrators, leading to fracture of thecomposite.

The impact results of the tested composites are given in Table 8.The variations of the impact toughness with fiber weight frac-

tion are depicted in Fig. 10. The maximum toughness is found forcomposites with 10 and 20 wt% of fibers. The incorporation of30% of fibers leads to a composite with a bad impact behavior, with18% less capability of absorbing energy considering the other fiberincorporations.

The variations of the maximum deflection with fiber wf are de-picted in Fig. 11. The addition of cellulosic fibers decreases themaximum deflection of the composites.

The incorporation of 30% of fibers leads to a composite withdecreasing on the deformation capabilities at break of around 33%.

4.4. Heat Deflection Temperature (HDT) measurement

The Heat Deflection Temperature (HDT) results are given inTable 9 and Fig. 12 for all fiber compositions.

As expected the incorporation of fibers increases the HDT. Themaximum synergetic effect is obtained with the incorporation of20% of fibers leading to an increasing of 15% on the HDT value.

4.5. Microscopy analysis

The previous mechanical characterization highlights the effectsof the adhesion between the polymeric matrix and the cellulosicfiber and of the dispersion of fibers.

The optical microscopy analysis images are shown in Fig. 13 andSEM analysis in Fig. 14. The incorporation of 30% wf of fiber leadsto a non-homogeneous composite and the incorporation until 20%wf drives to a homogeneous composite.

As possible to verify in Fig. 14, the matrix is a perfectly miscibleone. Is not possible to identify a PLA or a PHA phase in the matrix.

The SEM analysis also corroborates that the fiber dispersion ishomogenous until reach the 20% wf. After that the dispersion starts

to random and lost the homogeneity. That can be verified in Fig. 13(right) where it is possible to observe the flow induced orientation.

As seen in Fig. 15 the SEM analysis reveals that the fibers aredeboned of the matrix inducing a lower interfacial adhesion. Thisinterfacial behavior justifies the deviation of the experimental dataand the predicated values based into perfect adhesion.

5. Conclusions

The properties of biodegradable composites can be tailored toachieve a given performance.

Composites with a [30:70] [PHA:PLA] matrix and with a fibercontent of 10%, 20% and 30% (wf) were investigated in this work.The composites were injection molded, after being extruded andpelletized, and their mechanical (tensile, flexural and impact)and thermal (HDT) behaviors were assessed.

The increment of fiber amount increases the tensile and flexuralmoduli of the final composites. For the tensile modulus, a linearrelationship is found, following the modified Halpin–Tsai equation(or the rules of mixtures) denoting a homogeneous dispersion ofthe fibers for fiber incorporation until 20% (wf).

The ROM validates the flexural behavior determined by theexperimental tests. The experimental values allow us to concluded,again, that for fiber incorporation over than 20% (wf) the compositebecame non-homogeneous and therefore with reduced mechanicalproperties.

The moduli and maximum stress of the composites with fiberincorporation lower or equal to 20% (wf) can be estimated recur-ring to the presented prediction models.

The SEM analysis identifies a lower matrix/fiber interfacialadhesion that justifies the deviation of the experimental data andthe predicated values.

The incorporation of 20% wf fiber improves the impact energyabsorption and the Heat-Deflection Temperature.

Prediction models and material property characterizationallowed unambiguous detection of a maximum of fiberincorporation.

Once that the composites are from renewable sources thesedata gives an indication of the potential use of these compositesreplacing the petrol-based matrix composites with synthetic rein-forced fibers.

Acknowledgements

The authors acknowledge the Portuguese Science and Technol-ogy Foundation (FCT) for the financial support provided by the

N.C. Loureiro et al. / Composites: Part B 60 (2014) 603–611 611

project MIT-Pt/EDAM-SMS/0030/2008 – Assessment and Develop-ment of integrated Systems for Electric Vehicles. N.C. Loureiroacknowledges also the Portuguese Science and Technology Foun-dation (FCT) for the financial support provided by the Ph.D. GrantSFRH/BD/42978/2008.

References

[1] Holbery J, Houston D. Natural-fiber-reinforced polymer composites inautomotive applications. J Min Met Mater Soc 2006;58:80–6.

[2] Mohanty AK, Misra M, Drzal LT. Sustainable bio-composites from renewableresources: opportunities and challenges in the green materials World. J PolymEnviron 2002;10:19–26.

[3] Mohanty AK, Misra M, Hinrichsen G. Biofibres, biodegradable polymers andbiocomposites: an overview, vol. 276–277; 2000. p. 1–24.

[4] Fowler PA, Mark-Hughes J. Biocomposites: technology, environmentalcredentials and market forces. J Sci Food Agric 2006;86:1781–9.

[5] Pereira N, Sousa ML, Agnelli JAM, Mattoso LHC. Effect of processing on theproperties of polypropylene reinforced with short sisal fibers. In: Proceedingsof 4th international conference on wood fiber plastic composites, Madison,United States of America; 1997.

[6] Mattoso LHC, Ferreira FC, Curvelo AAS, Lignocellulosic-plasticcomposites, SINCdo Brazil, São Paulo, Brazil; 1997.

[7] Oksman K. Improved interaction between wood and synthetic polymers inwood/polymer composites. Wood Sci Technol 1996;30:197–205.

[8] Osman MA, Atallah A, Müller M, Suter UW. Reinforcement ofpoly(dimethylsiloxane) networks by mica flakes. Polymer 2001;42:6545–56.

[9] Eichhorn SJ, Baillie CA, Zafeiropoulos N, Mwaikambo LY, Ansell MP, Dufresne A,et al. Review: current international research into cellulosic fibers andcomposites. J Mater Sci 2001;36:2107–31.

[10] Loureiro NC, Esteves JL, Viana JC. Caracterização Mecânica de Misturas de PLA/ABS. In: Proceedings of ENMEC2010 – Encontro Nacional de Materiais eEstruturas Compósitas, Porto, Portugal; 2010.

[11] Auras R, Harte B, Selke S. An overview of polylactides as packaging materials.Macromol Biosci 2004;4:835–64.

[12] Loureiro NC, Esteves JL, Viana JC. Mechanical characterization of PLA/PHAblends. In: ICCS16 - 16th international conference on composite structures,Porto, Portugal; 2011.

[13] Gérard T, Budtova T. PLA-PHA Blends: morphology, thermal and mechanicalproperties. In: Proceedings of BIOPOL’2011 – international conference onbiodegradable and biobased polymers; 2011.

[14] Sathyanarayana KG, Sukumaran K, Mukherjee PS, Pavithran C, Pillai SGK.Natural fibre-polymer composites. Cem Concr Compos 1990;12:117–36.

[15] Paul MA, Alexandre M, Degee P, Henrist C, Rulmont A, Dubois P. Newnanocomposite materials based on plasticized poly(L-lactide) and organo-modified montmorillonites: thermal and morphological study. Polymer2003;44:443–50.

[16] Mathew AP, Dufresne A. Morphological investigation of nanocomposites fromsorbitol plasticized starch and tunicin whiskers. Biomacromolecules2002;3:609–17.

[17] Lundquist L, Marque B, Hagstrand P-O, Leterrier Y, Manson J-AE. Novel pulpfibre reinforced thermoplastic composites. Compos Sci Technol2003;63:137–52.

[18] Mathew AP, Oksman K, Sain M. Mechanical properties of biodegradablecomposites from poly lactic acid (PLA) and microcrystalline cellulose (MCC). JAppl Polym Sci 2005;97:2014–25.

[19] Guimarães D. Efeito das condições de injecção nas propriedades de PLAreforçado com fibras celulósicas. Master Thesis in Polymer Engineering,University of Minho, Portugal; 2009.

[20] Mishra S, Tripathy SS, Mishra M, Mohanty AK, Nayak SK. Novel eco-friendlybiocomposites: biofiber reinforced biodegradable polyester amide composites– fabrication and properties evaluation. J Reinf Plast Compos 2002;21:55–70.

[21] Loureiro NC, Esteves JL, Viana JC, Ghosh S. Development and characterizationof renewable source composites to replace petrol-based polymers into interiordoor trims. In: Proceedings of ECCM15 – 15th European conference oncomposite materials, Venice, Italy; 2012.

[22] Kalaprasad G, Joseph K, Thomas S, Pavithran C. Theroetical modeling of tensileproperties of short sisal fibre-reinforced low-density polyethylene composites.J Mater Sci 1997;32:4261–7.

[23] Ahmed S, Jones FR. A review of particulate reinforcement theories for polymercomposites. J Mater Sci 1990;25:4933–42.

[24] Savastano Jr H, Warden PG, Coutts RSP. Brazilian waste fibres as reinforcementfor cement-based composites. Cem Concr Compos 2000;22:379–84.

[25] Agopyan V, Savastano Jr H, John VM, Cincotto MA. Developments on vegetablefibre-cement based materials in São Paulo, Brazil: an overview. Cem ConcrCompos 2005;27:527–36.