Pdn.sciencedirect.com Science Ob

8/2/2019 Pdn.sciencedirect.com Science Ob

1/11

Available onl ine at www.sciencedirect.comBCIINCI@DRICT COMPOSITESSCIENCE ANDTECHNOWGY

ELSEVIER Composites Science and Technology 65 (2005) 1582-1592www.elsevier.comllocate/compscitech

Mechanical properties of short flax fibrebundle/polypropylene composites: Influence ofmatrix/fibre modification, fibre content, water uptake and recyclingA. Arbelaiz, B. Fernandez, I.A. Ramos, A. Retegi, R. Llano-Ponte, 1. Mondragon *

Materials + Technologies Group, Dpto. Ingenieria Qutmica y M. Ambiente, Escuela Univ Politecnica, Euskal HerrikoUnibertsitatealUniversidad del Pals Vasco, Plaza Europa 1,20018 Donostia-San Sebastian, Spain

Received 10 July 2004; received in revised form 19 January 2005; accepted 27 January 2005Available online 23 February 2005

AbstractThe aim of this work was to compare the effect of modification way of short flax fibre bundle/polypropylene (PP) composites on

mechanical properties. Modification was carried out on fibre surface and also modifying PP matrix using several amounts of maleicanhydride-polypropylene copolymer (MAPP) as compatibilizer. The optimum doses of two different MAPP compatibilizers havebeen obtained. The effect of fibre bundle loading on composite mechanical properties was also analysed. The influence of wateruptake on the sorption characteristics of composites has been studied by immersion in distilled water at room temperature. Theeffects of fibre bundle loading and also the use of MAPP modification on both water sorption and mechanical properties were alsoevaluated. Results showed that using MAPP as coupling agent, mechanical properties of composites improved, and water uptakerate clearly decreased. However, after long period of water immersion mechanical properties drastically decreased. On the otherhand, mechanical recycling of flax fibre bundle/PP composites has been shown to be feasible. 2005 Elsevier Ltd. All rights reserved.Keywords: Flax; A. Fibres; A. Polymer-matrix composites; B. Environmental degradation; B. Mechanical properties

1. IntroductionNowadays, ecological concern has resulted in a re-

newed interest in natural materials and issues such asrecyclability and environmental safety are becomingincreasingly important for the introduction of newmaterials and products [1]. Lignocellulosic fibres havemany advantages as they are biodegradable, renewableand environmentally friendly. They have acceptable spe-cific properties and comparing to glass fibres, theyreduce dermal, respiratory irritation during handlingas well as tool wear. However, the main disadvantageof natural fibres is their hydrophilic nature that lowers* Corresponding author. Tel.: +34943017271; fax: +34943017200.E-mail address:[email protected] (1 . Mondragon).

0266-3538/$ - see front matter 2005 Elsevier Ltd. All rights reserved.doi: 10.10 16/j.compscitech.2005.0 1.008

the compatibility with hydrophobic polymeric matricesduring composite fabrication. They also present poorenvironmental and dimensional stability [1-5]. Amor-phous cellulose and hemicellulosesare mostly responsi-ble for the high water uptake of natural fibres, sincethey contain numerous easy accessiblehydroxyl groupswhich give strong hydrophilic character to fibres.Due to hydrophilic character, swelling by water up-

take can lead to microcracking of the composite anddegradation of mechanical properties [1]. Therefore,water uptake is one of the most serious problems thatprevents a wider use of natural fibre composites [6-9J,in fact, in wet conditions, mechanical performance ofcomposites decreases [6,7,9]. The possibility for usingthese materials in outdoor applications makes it neces-sary to analyse their mechanical behaviour under the
http://www.sciencedirect.com/http://www.elsevier.comllocate/compscitechmailto:address:[email protected]:address:[email protected]://www.elsevier.comllocate/compscitechhttp://www.sciencedirect.com/


2/11

A. Arbelaiz et al. / Composites Science and Technology 65 (2005) 1582-1592

influence of the weathering action in the long term [6].The generation of a stronger interface between matrixand reinforcement material could reduce the hygroscop-icity of lignocellulosic-based materials [9].In order to improve adhesion between cellulosic fibresand polypropylene (PP) matrix, coupling or compatibi-lizing agents can be used.Graft copolymers of PP and maleic anhydride areknown to be very effective additives for lignocellulosic/PP composites [5,10-13]. The effectiveness can be dueto a better compatibility between fibres and PP matrix[14-16J as also to a better dispersion of flax fibres inPP. In the literature survey [8,17-21J different compati-bilizers have been used for lignocellulosic fibre/PP com-posites. In a previous work [14Jthe influence of differentflax fibre surface modifications on fibre mechanical andflax/PP interface properties has been reported. Resultsshowed that MAPP treated fibre presented the best per-formances. MAPP treatment could also reduce hydro-gen bonding that tends to bind fibres together [15].The formation of covalent bonds between OH groupsof cellulose and anhydride groups of MAPP [5,22J,which can be confirmed by IR studies [11,23,24J, isone of the reasons for strength improvement. On theother hand, entanglements between MAPP and PPchains can be created, so acting as physical cross-links[13,16]. Therefore, when stresses are applied to PP ma-trix, they can be transmitted to the fibre through thesephysical cross-links.Two properties of maleated PPs that could influencetheir effectiveness as coupling agents for natural fibre/PPcomposites are molecular weight, which affects entangle-ment with the matrix chains, and acid number, whichdetermines the functionality present in the couplingagent. To develop sufficient stress transfer properties be-tween the matrix and the fibre, MA groups in MAPPshould interact or even react with the OH groups onthe fibre surface and the polymer chains of the MAPPshould be long enough to permit entanglements withthe PP in the interphase [13].In this work two maleated coupling agents with dif-ferent acid numbers and molecular weights were usedfor determining the effect of these parameters on cou-plen effectiveness. The objective of this work was tocompare the influence of both fibre surface modificationand matrix modification on mechanical properties andalso to determine the optimum dose of each MAPPcompatibilizer in flax fibre bundle/PP composites. Afterdefining the optimum MAPP dose, mechanical proper-ties of composites at different fibre bundle loadings werecompared. On the other hand, the environmental behav-iour of both unmodified and MAPP-modified flax fibrebundle/PP composites was evaluated by comparingwater uptake and tensile properties. Finally, composites

based on MAPP-modified flax fibre bundle/PP compos-ites were grounded and reinjected in order to know the

1583

influence of number of passings through injectionmoulding machine on tensile properties.

2. Experimental2.1. Materials

A commercially available PP "Eltex-P HV200"produced by Solvay with a Melt Flow Index oflu g/In min (at 230C and 2.16 kg) was used as poly-meric matrix. The reinforcement was a natural flax fibrebundle obtained by a retting process, which consists inthe biological action of bacteria in an aqueous mediumwhere waxes and pectins are removed. These fibre bun-dles were kindly supplied by Finflax (Finland). Flaxfibre bundles were first chopped to a length of approxi-mately 30 mm and the diameter values varied from 10 to120 urn. MAPP, Epolene E43 and G3003, kindly sup-plied by Eastman Chemical were used as couplingagents. Epolene E43 has a low molecular weight(M,=3.900, Mw=9.100),0.934 g/mL density and acidnumber of 45. On the other hand, Epolene G3003 hashigher molecular weight (M; =27.200, Mw=52.000),0.912 g/mL density but lower acid number, 8, thanE43 one [25].2.2. Surface modification

Fibre bundles were treated with two differentamounts of Epolene E43, 5 and 10wt%. Treatment pro-cedure was reported in a previous work [24].2.3. Matrix modification

PP matrix was modified with MAPP. For both typesof MAPP, 1,2, 3,4, 5, 10 and 20 wt% amounts respectto fibre bundle content were used. The amount ofMAPP was fed directly into the melt mixer.

2.4. Compounding and processingCompounding was carried out using a melt mixer(Haake Rheomix 600 with two Banbury rotors). Themixing temperature was set at 180C. First PP andMAPP pellets were charged and after melting dried fibrebundles were added. The systems were mixed during5min at 40 rpm. The loading of flax fibres varied from

o to 60 wt%. A percent of 30 wt% of fibre bundle wasused to compare the effect of amount and type of MAPPcoupling agent on mechanical properties. The mixturewas pelletized and kept in a vacuum oven at 100Cfor 12h. The moulding of the dried pellets was carriedout in an injection-moulding machine (Battenfeld Plus250). Samples were moulded for tensile and flexural tests


3/11

1584 A. Arbelaiz et al. / Composites Science and Technology 65 (2005) 1582-1592

3. Results and discussionccording to ASTM D638 and ASTM D790M stan-dards, respectively.2.5. Mechanical testing of composites

Tensile and three-point bending tests were carried outusing an universal mechanical testing machine Instron,model 4206. The crosshead speeds used were 5 and1.7mm/min for tensile and flexural tests, respectively.The gauge length used for tensile test was 115mm andthe span for flexural test was 64 mm. At least five spec-imens were tested for each set of samples and the meanvalues are reported. On the other hand, impact test werecarried out using a Rosand instrumented falling-weightimpact analysis system with a "Charpy" specimen sup-port. Following testing conditions have been used: spanlength was 60 mm, drop mass and drop rate were 0.7 kgand 3.29 mis, respectively, and unnotched flexural spec-imens with 10x 4 x 80 mm' dimensions were used. Dur-ing the impact, the resistive force exerted by the sampleon the striker was measured as a function of time, andstored for subsequent display and analysis. The impactenergy measured only can be used for relative compari-son as it does not give the accurate toughness of thematerial.

2.6. Fibre length and diameter measurements

Fibre length and diameter were analyzed in injectionmoulded specimens to investigate the effect of couplingagent and fibre content. Matrix polymer was dissolvedin hot xylene and a Nikon Eclipse E600W microscopywith software analySIS of Soft Imaging System wasused for fibre length and diameter measurements. Ammimum of 300 measurements were tested for eachsample.2.7. Ageing of samples

Five composite samples, with ASTM D638 standarddimensions, were immersed in distilled water at roomtemperature to study the kinetics of water uptake.Weight increase due to sorbed water was periodicallymeasured. When samples were taken out from water,the wet surface of the sample was quickly dried,weighed, and then samples were again immersed inwater. This process continued during 7 months. Afterthis period, samples were mechanically tested. The wateruptake at time t was calculated from Eq. (1) [26,27Jj},m(t) =( W I ; ' o W o ) * 100, (1)where Wo is the weight of dry specimen and Wt is theweight of wet specimen at t time. Water mass incrementat equilibrium (infinite time) is represented by j},m [oo).

3.1. Matrix modification versus fibre modification

Figs. lea) and (b) show the effect of modification wayon both tensile and flexural properties for 30 wt% flaxfibre/PP composites. When composites prepared via fi-bre modification or via matrix modification with MAPPare compared, both ways caused a significant increase instrength, which is agreement with other author results[12].MAPP might act as a dispersing agent between po-lar fibres and apolar matrix resulting in better dispersionof the fibres [28]. Irrespective to modification way used,strength values obtained were similar. Surface modifiedcomposites show slightly higher modulus than matrixmodified ones. Results have to be taken carefully assmall differences in fibre content could influence them[29]. Felix and Gatenholm [11J found for cellulose/PPcomposites that cellulose fibre surface treatment withMAPP increased tensile strength by 80% at a fibre load-ing of 40 wt%. Karmaker and Youngquist [5Jfound forjute/PP composites that addition of 3 wt% of MAPP tothe kinetic mixer at a fibre loading of 50 wt% increasedtensile strength by 100%. Thus, both modification waysare valid to get a composite with higher mechanicalproperties than unmodified ones. However, fibre surface

I D Fiber modification ' I (a)o Ma tr ix mod if ic a ti on6 Untreated 300040 :0:Q

C O mIl.. 2500 :.. .. ..~ s : :~ 30 "Utt * ~200020 ~

1500Untreated 5wt% MAPP 10wt%MAPP

70

ID Fiber modification I (b)o Ma tr ix mod if ic a ti o n6. Untreated 6000

60 rrC ? mIl.. II 5000~~ "Utl- ~

50 :a:t 1 : 4000*. . .40 Untreated 5wt% MAPP 10wt%MAPP

Fig. 1. Mechanical behaviour of 30wt% flax fibre/PP composites as afunction of modification way: (a) tensile; (b) flexural; strength (opensymbols) and modulus (solid symbols).


4/11


modification with MAPP would probably be expensivesince it involves solvent utilization, and therefore itwould not be industrially practicable [12]. So that, thefollowing results correspond to MAPP-modified PPmatrix composites.3.2. Matrix modification with MAPP

Figs. 2(a) and (b) show the effect of MAPP type ontensile and flexural properties as a function of theamount of coupling agent used. Both MAPP lead tosimilar trends in composites strength. Both tensile andflexural strength increased with increasing MAPP con-tent until a plateau region was reached. It is worth not-ing that, 5 and 10wt% compatibilizer for E43 andG3003, respectively, were the optimum doses, as flexuralstrength increased about 40% and 60%, respectively. Fortensile strength, the maximum improvement for E43 andG3003 were 42% and 58%, respectively. After addingonly 1wt% of E43 compatibilizer, composites strengthclearly increased. This improvement is due to the en-hanced stress transfer from the matrix to the fibre viathe compatibilizer [10,22]. Further increase in the com-patibilizer beyond 1wt% of both types of MAPP in-

(a): 2 : 3000:0:

2500 m~-u~I 20001:1500

0 5 10 15 20Coupling agent (wt %)

Q(b) 6000

40

c o0...63010-

20

60c o0...2:10-

50

oS -40

0

m5000-'5:

-u~4000

5 10 15 20Coupling agent (w t %)

Fig. 2. Influence of MAPP type and amount on mechanical behaviourof 30 wt% flax fibre/PP composites: (a) tensile; (b) flexural; strength(open symbols) and modulus (solid symbols).

1585

creased composites strength value until a maximum.Comparing strength values of modified composites,E43 seems to be more effective until 3 wt% of couplingagent whereas beyond this percentage, G3003-modifiedcomposites showed better strength performance. As anoverall trend, when MAPP was added modulus valuesslightly increased. This improvement could be relatedto better dispersion of fibre bundles on PP matrix. How-ever, at high coupling agent concentrations (20 wt%)modulus value showed a small drop.As a matter of fact, improvements on mechanicalproperties of composites are significantly dependent onthe grafting rate and on the average molar mass of thegraft copolymer coupling agent [30,31]. The behaviourof E43 and G3003-modified composites can be ex-plained taking in account both modifier acid numberand molecular weight. Acid number measures the func-tionality present in the coupling agent, being dependentupon MA units grafted per polymeric chain. Molecularweight is related to the ability to create entanglementsbetween coupling agent and PP matrix [13,22]. G3003has a lower amount of maleic groups per chain length,so at low modifier content it has not enough maleicgroups to produce an optimal coupling efficiency. Be-yond 3wt% of coupling agent, G3003 has enoughmaleic groups to create interactions with flax fibre bun-dle and thus a better stress transfer from the matrix tothe fibre bundle is expected. E43 has lower molecularweight, so polymer chains are shorter than G3003 ones,being the chance of entanglements with matrix lowerthan for G3003 one. Felix et al. [30Jobserved for cellu-lose/PP composites that using two different MAPP, withlow and high molecular weight, the longer the PP chainof the modifier the better tensile properties were. Theysuggested that high molecular weight MAPP has moreflexible PP chains which are able to diffuse deeper intothe matrix. Thus, MAPP chains become more involvedin inter-chain entanglements and thereby they contrib-ute to the mechanical continuity of the system.Fig. 3 shows the impact energy at failure of compos-ites as a function of coupling agent content. Impactresistance decreased drastically when fibre bundles wereadded. Besides, composites modified with both MAPPshowed similar trends as impact resistance increasedupon modifier content up to a maximum was achieved.A reduction of impact strength was observed for bothmodifiers at 20 wt%, which can be attributed to themigration of a high amount of compatibilizer from thefibres to the matrix, causing self-entanglement amongthe compatibilizer chains rather than with the polymermatrix, thus resulting in slippage [10,22].The maximumimpact resistance for E43-modified composites wasfound between 5 and 10wt% whilst for G3003-modifiedcomposites was at 10wt%. These optimum values aresimilar to those obtained for flexural and tensilestrengths. Therefore, it can be concluded that for


5/11

70


NE-- .--,~ 608!:::l$1 0>-e > 20Q)cQ)g 10a.E

~E43 3 1

,~~G3003c:::J PP

r J ~ ~~ ~ ;J;~~

~ ;I;; ~ ~ ~ ~~ ~ ~~ ~ ~~ ~ ~~ ~ ~ ~o o 1 2 3 4 5Coupling agent (w t %)

10 20

Fig. 3. Influence of MAPP type and amount on impact energy ofcomposites with 30wt% flax fibre.

different MAPP coupling agents, there is an optimumdose which results in the best improvement on mechan-ical behaviour of flax fibre bundle/PP composites. Sim-ilar observations were made by Qiu et al. [32J forfibrous cellulose/PP composites. They observed that ten-sile strength highly increased with an increase in theMAPP content, showing a maximum around 10wt% be-yond which tensile property decreased. They suggestedthat there is a critical amount of compatibilizer at whichMAPP exhibits the strongest interactions with cellulosefibres as well as with PP matrix, which is similar to thatobserved in our work. Keener et al. [33Jfound for ligno-cellulosic fibre/PP composites, with 30 wt% fibre con-tent, that 3% of maleated PP Epolene G30l5, whichacid number and molecular weight are intermediate pro-vided optimal performance. Mechanical results seem toindicate that coupling of MAPP anhydride groups onflax fibre surface, but also a higher extent of physicalentanglements with PP matrix because of increase onMAPP molecular weight, do have an important contri-bution on fracture toughness.All results reported below are related to the optimaldoses of MAPP.3.3. The effect of fibre bundle loading

Figs. 4(a and b) and 5(a and b) show both tensileand flexural properties of composites as a function offibre bundle loading and MAPP type. For unmodifiedcomposites tensile strength decreased as fibre bundlecontent increased whereas it showed similar flexuralstrength values than for neat PP. This is in agreementwith the trend observed in other unmodified lignocellu-losic fibre-reinforced composites [10-12,27,32]. Thisbehaviour suggests that there was small stress transferfrom the matrix to the fibres irrespective of the amountof fibre present [10,22].When MAPP was added, mod-

60r~========~------------~--1 (a)50

~ Unmodified 5wt%E43_ 10 w t%G3003DPP

40c o0...~ 30

20

10

10 20 30 40Fibre (wt %)

600

3000

~ Unmodified 5wt%E43a0 wt% G3003DPP

1000

10 20 30 40Fibre (wt %) 600

Fig. 4. Tensile behaviour as a function of fibre loading and MAPPtype: (a) strength; (b) modulus.

ified composites showed higher strength values thanthose for unmodified ones and neat PP as MAPP addi-tion increased fibre/matrix adhesion [14,16,34,35]. Thehighest strength improvement was obtained whenG3003 coupling agent was used. The effect of MAPPmodification became more pronounced with an in-crease of fibre bundle content [10-12]. At a fibre bun-dle loading of 60 wt%, the flexural strength increasedby 93% and 118% for E43 and G3003-modified com-posites, respectively. In the case of a 20 wt% fibre bun-dle content the increase was 26% and 27% for E43 andG3003-modified composites, respectively. Strength ofMAPP-modified systems increased with the amountof fibre bundle present, probably as a consequence ofboth improved fibre bundle dispersion in the matrixand better fibre/matrix adhesion, which is in agreementwith other works [10,36].Addition of MAPP, as mentioned above, slightly in-creased modulus values. However, other authors re-ported [5,11,22,37J that coupling agent addition hasnot significant effect on tensile and flexural modulus.As expected, increasing of fibre bundle content resultedin higher modulus values. Similar observations were re-ported by other authors [11,22,37Jfor other lignocellu-losic based PP composites.


6/11


100~ Unmodified (a)8wt%E43

80 a0 wt % G3003DPP60C ?o,~ 401::)-20

0 10 20 30 40 50 60Fibre (wt %)

9000 ~ Unmodified8wt%E43a0 wt % G3003DPP6000

C ?o,~W-

3000

10 20 30 40Fibre (wt%)

50 60

Fig. 5. Flexural behaviour as a function of fibre loading and MAPPtype: (a) strength; (b) modulus.

70NE

~

-- .--,~~ 60:::l$1 0>-e >Q)c 20)

1 :5ella. 10E

020 30

~ Unmodified~ 5wt%E43~ 10wt%G3003c::::J pp

40Fibre (w t %)

50 60

Fig. 6. Impact energy as a function of fibre loading and MAPP type.

Fig. 6 shows impact energy as a function of fibre bun-dle loading and MAPP type. All composites showedlower impact energy than neat PP because the additionof the fibre bundles creates regions of stress concentra-tions that require less energy to initiate cracking [10].The impact energy of unmodified composites decreasedwhen increasing fibre bundle content. However, modi-

1587

fied composites showed a maximum, thereafter decreas-ing. For E43 and G3003-modified composites, themaximum impact energy was obtained at 30 and40 wt% fibre bundle contents, respectively. Thus,improving fibre/matrix adhesion and fibre bundle dis-persion through use of MAPP increases the resistanceto crack initiation at the fibre/matrix interface, and thefall in impact strength with the addition of fibre bundlesis not as dramatic [10]. As for tensile and flexuralstrength, G3003-modified composites showed the high-est impact energy due to the important contributionon impact behaviour of the higher amount of entangle-ments with respect to the E43-modified ones.3.4. Fibre breakage analysis

Processing affects the final fibre aspect ratio and thusthe mechanical properties of the product [38].Processingtechniques such as internal mixing and injection mould-ing can cause a high fibre attrition [5,39J owing to thestrong shear stresses acting in the viscous molten poly-mer [40]. Fibres can be broken into smaller fibres ifthe hydrodynamic forces exceed the fibre strength and/or the cohesive forces between the individual fibres[39]. Natural fibres have weak links as natural and arti-ficial flaws as well as kink bands [14]. These weak linksare the most probable rupture points along the fibrelength when the fibre is mechanically stressed [5,40].The final diameter (d), fibre length (L) and aspect ra-tio (LId) values of the fibres extracted from compositesare reported in Table 1. After injection molding, fibrediameter decreased and for all system the final fibrediameter is around 20 urn, which means that the used fi-bre bundles have been separated through processing.Besides, the original fibre length drastically decreasedas higher fibre content was. This fact led to a reductionof around 30% in the aspect ratio for all compositeswhen the fibre content was raised from 20 to 60 wt%.Joseph et al. [41Jfound for sisal/PP +MAPP com-posites that viscosity increases with fibre loading. There-fore higher melt viscosity resulted in a reduction ofaspect ratio. Using MAPP the average fibre length is sli-gthly higher than that for unmodified composites for allfibre contents. One of the reasons may be that MAPPhas a low molecular weight compared to PP matrix, soMAPP could act as a lubricating agent. Thereby areduction of shearing among fibres and molten PP couldoccur [5]. The slight higher aspect ratios but above allthe improvement of fibre/matrix adhesion provide aconsiderable reinforcing effect, as discussed above.3.5. Water uptake

Joly et al. [34,42Jfound for different MAPP that thegrafting extent corresponding to covalently bondedcompatibilizing agent on the cellulose fibre (degree of


7/11

1588 A. Arbelaiz et al. I Composites Science and Technology 65 (2005) 1582-1592

Table 1Fibre characteristics after processing as a function of fibre content and coupling agentFibre content (wt%) Unmodified E43 G3003

d (um) L (um) Lid d (um) L (um) Lid d (um) L (um) Lid20 19 9 158 75 8.3 20 10 197 105 9.9 25 10 236 108 9.430 18 7 154 77 8.6 20 7 203 109 10 21 5 194 96 9.240 20 8 146 76 7.3 18 7 155 68 8.6 22 8 155 73 7.050 18 7 135 69 7.5 18 6 150 67 8.3 18 6 147 65 8.260 19 8 118 59 6.2 20 8 141 68 7.0 20 6 122 43 6.1

crystallinity was 65%) was very low (~l% or less), onlycelluloses one OR group out of 19.000 to 45.000 reactedwith MAPP. So that slight differences in water uptakeshould exist between unmodified and MAPP-modifiedcomposites.Taking into account both acid number and amountof compatibilizer used in MAPP modified composites,it is clear that E43 modified composites should presentmore chance to block more fibre hydroxyl groups thanG3003 modified composites ones, thus being a bettercandidate for decreasing water uptake. Therefore thestudy on water uptake has been performed only forE43-modified composites.

20 (a)

15

~1 0E


8/11


considered as rectangular with 10 and 150mm widthand length, respectively. Diffusivity and equilibrium up-take values with different fibre bundle loadings forunmodified and 5 wt% E43-modified composites are gi-ven in Tables 2 and 3, respectively. The results for thediffusion coefficient are similar to the values reportedby other authors [51,53]. Use of MAPP remarkably re-duced diffusion coefficient, being higher when fibre bun-dle loading was increased, whilst water uptake atequilibrium slightly decreased. Similar observationswere found by George et al. [44J for pineapple-leaf fi-bre/low density polyethylene composites. Peijs et al. [1Jfound that the use of MAPP lowers the diffusivity forflax fibre mat in combination with PP matrix. For im-proved adhesion between matrix and fibres, rate of thediffusional processes decreases since there are fewer gapsin the interfacial region and also more hydrophilicgroups as hydroxyls are blocked by the coupling effect[53].Use of MAPP slightly reduced equilibrium water up-take for MAPP-modified samples, this small reductionbeing attributed to the improved interfacial adhesion[9,26,44Jthat avoid an easy penetration of water mole-cules into the modified composites [9Jand reduced wateraccumulation in the interfacial voids which preventswater from entering the natural fibre [26]. Rana et al.[22Jmentioned that the decrease of water uptake usingMAPP might be attributed to some of fibre hydrophilic-OR groups reacting with maleic anhydride to form es-ter linkages and thereby giving slightly lower water-uptake values.

Table 2Equilibrium uptake, diffusion coefficient and corrected diffusioncoefficient values with different fibre bundle loadings for unmodifiedcompositesFibre (wt%) fl.m (00) (%) D * 109 (cm2js) D; * 109 (cm2js)30 5.92 1.52 0.7540 9.09 4.73 2.3250 11.36 9.72 4.7860 13.96 17.45 8.57

Table 3Equilibrium uptake, diffusion coefficient and corrected diffusioncoefficient with different fibre bundle loadings for 5 wt% E43 modifiedcompositesFibre (wt%) fl.m (00) (%)30405060

8.5310.5613.08

1.883.9511.8

0.921.945.80

1589

3.6. Effect of water immersion on mechanical propertiesAfter aging in distilled water at room temperature foraround 7 months, tensile properties of all types of sam-

ples show significant drop (Fig. 8(a) and (b)). Mechani-cal properties decreased because water moleculeschanged the structure and properties of fibres, matrixand interface between them. When fibre/matrix interfaceis accessible to moisture from the environment, the cel-lulosic fibres tend to swell, thereby developing shearstresses at the interface, which favours ultimate debond-ing of the fibres, which in turn causes a reduction in ten-sile strength [51].It is also seen that tensile properties forMAPP-modified systems were better than for unmodi-fied ones. Similar observations for bamboo-glass rein-forced PP composites were observed by Thwe andLiao [26]. In contrast, for Kenaf/PP composites, Caul-field et al. [54Jfound that after water exposure, no cleardistinction was observed between unmodified andMAPP-modified samples. George et al. [44Jinvestigatedthe variation of tensile strength for pineapple leaf fibre/LDPE composites after water immersion for differentperiods of time, finding that as immersion time in-creased, composite tensile strength decreased. Singh

60 I C Before ageing I 3000o After aqeino50 I !

C O ! ! 2000 m0 . . . 40 ~6 -utl- :I ~# :t 100030 ~ # ~ : m :(a) ~ i0 00 10 20 30 40 50 60Fibre (wt%)

60 I C Before ageing I ! 3000o After ageingI

50C O I I rr 2000 m0 . . . :0: -6 40 :c: ~tl- I: :t -u :I ~: t : I = :0: 100030 :0: :e : :0: :II

20 (b) 00 10 20 30 40 50 60Fibre (wt %)

Fig. 8. Variation of tensile strength (open symbols) and modulus(solid symbols) of composites with different fibre contents before andafter ageing in distilled water: (a) unmodified; (b) 5 wt% E43-modified.


9/11


et al. [55Jfound that the reduction in strength for jute/phenolic composites with increasing humidity levels isexpected to depend on the amount of moisture/waterwhich disturbs the mechanical integrity of compositesby affecting the matrix, the fibres and the fibre/matrixinterface simultaneously.After water immersion, the stress-strain behaviour forcomposites containing different fibre loadings was exam-ined (Fig. 9). Curves extended into the nonlinear region inall cases. Modulus, which was determined from the initialslope of the stress-strain curve, was similar for both fibreloadings. This trend is in contrast with results obtainedpreviously with composites before ageing, suggesting thatafter water immersion composites fibres degraded andsubsequently composites modulus was irrespective to fi-bre content. Sreekala and Thomas [56Jobserved for oilpalm fibres that in the swollen stage the strength and mod-ulus were found to decline. Joseph et al. [5lJ observed thatafter exposure sisal/PP composites in boiling water for7 h, composites modulus decreased due to plasticizationeffect. That is, the absorbed water molecule reduces theintermolecular hydrogen bonding between cellulose mol-ecules in the fibre and establishes intermolecular hydro-gen bonding between cellulose molecules in the fibre andwater molecules, thereby reducing the interfacial adhe-sion between the fibre and the matrix, and as a resultdecreasing composites mechanical properties. Espertet al. [53Jobserved that after water absorption oflignocel-lulosic/PP composites, SEM micrographs clearly showedthe degradation of the fibres and the loss of adhesion be-tween fibre and matrix. Stamboulis et al. [7Jconcludedthat rather than the result of interface failure or matrixcracking, the mechanical degradation of the compositesat relatively long immersion times ismost likely the resultof fibre degradation.Although after water immersion tensile strengthdrop, composites containing MAPP still showed higherstrength.

30

c o 200..61:)-

10

3 6E(%)

Fig. 9. Tensile stress vs. deformation curves for unmodified (solidsymbols) and 5 wt% E43-modified (open symbols) composites with 20and 60 wt% flax fibres after water immersion.

(a)60 ~

c o 400..~tl-20

oFibre (wt %)

6000 ,....--------------------, (b)

LlJ

o 10 20 30 40Fibre (wt%)

50 60

Fig. 10. Tensile behaviour of 5 wt% E43-modified composites as afunction of fibre loading and number of passings: (a) strength;(b) modulus.

3.7. Mechanical recyclingFigs. lO(a) and (b) show the effect of number of pas-sings through injection moulding machine on tensileproperties at different fibre bundle loadings. After pass-ing 4 times, tensile properties only showed a small de-crease. A similar trend was shown by Joseph et al. [57Jfor 20 wt% sisal fibre/LDPE matrix composites. It isworth noting that after passing 3 times mechanicallyrecycled composites through injection moulding ma-chine, tensile properties were similar to those for unrecy-cled materials. So that, mechanical recycling of flax fibrebundle/PP composites is possible as mechanical proper-ties only slightly changed.

4. Conclusions

9The influence of MAPP amount and type on mechan-ical properties of flax fibre bundle/PP composites hasbeen investigated. The effects of MAPP addition onboth water uptake and rate have also been studied.

Use of MAPP as coupling agent improved mechanicalproperties of composites by enhancing the adhesion


10/11


between flax fibre bundles and PP. Besides, fibre bundlesurface treatment with MAPP or matrix modificationwith MAPP had similar effects on mechanical propertiesof composites.Both acid number and molecular weight of MAPPare important parameters in order to determine its effec-tiveness. Results suggest that there is a critical amountof compatibilizer, dependent upon its constitution andmolecular weight, at which the strongest interactions be-tween lignocellulosic fibres and PP matrix occur.Mechanical properties showed that 5 and 10wt% com-patibilizer are the optimum doses for E43 and G3003,respectively. Besides, the improvement in mechanicalproperties was higher for composites modified withG3003 than for those with E43 since G3003 highermolecular weight can lead to more entanglements be-tween MAPP chains and PP matrix.Stiffness of unmodified and MAPP-modified compos-ites increased as fibre bundle content higher was, beingMAPP-modified composite modulus slightly higher.However, the strength for unmodified flax fibre bun-dle/PP composites decreased with flax fibre bundle con-tent, as a consequence of a poor interfacial adhesion,while for MAPP-modified composites the opposite trendwas observed.Water uptake increased with fibre bundle content.MAPP-modified composites exhibited a remarkablereduction in water uptake rate. Besides, due to a betterinterfacial bonding between flax fibre bundle and PP

matrix, the equilibrium water uptake showed a smallreduction. Although the use of MAPP slightly decreasescomposite affinity to water, long period of water immer-sion degrades fibres mechanical properties and subse-quently composites properties. Therefore, flax fibrebundle/PP composites should not be used in applica-tions where water uptake is of critical importance.After passing MAPP-modified flax fibre bundle/PPcomposites several times through injection mouldingmachine, mechanical properties only slightly changed.So that, mechanical recycling of flax fibre bundle/PPcomposites is one feasible option to re-use this type ofmaterial.

AcknowledgementThis research was performed as a part of the FP5GROWTH-ECOFINA G5RD-CTl999-00l47 projectin the framework of the 5th European Research Pro-gram (www.ecofina.org).

References[1] Peijs T, Van Melick HGH, Garkhail SK, Pott GT, Baillie CA.

Natural-fibre-mat reinforced thermoplastics based on upgraded

1591

flax fibres for improved moisture resistance. In: ECCM-8 confer-ence, Naples (Italy); 1998. p. 119-26.

[2] Brouwer WD. Natural fibre composites: Where can flax competewith glass? SAMPE J 2000;36:18-23.

[3] Mohanty AK, Misra M, Hinrichsen G. Biofibres, biodegradablepolymers and biocomposites: an overview. Macromol Mater Eng2000;276/277: 1-24.

[4] De Bruijn SJCM. Natural fibre mat thermoplastic productsfrom a processor's point of view. Appl Comp Mater2000;7:415-20.

[5] Karmaker AC, Youngquist JA. Injection moulding of polypro-pylene reinforced with short jute fibres. J Appl Polym Sci1996;62: 1147-51.

[6] Cantero G, Arbelaiz A, Mugika F, Valea A, Mondragon I.Mechanical behaviour of wood/polypropylene composites: effectsof fibre treatments and ageing processes. J Reinf Plast Comp2003;22:37-50.

[7] Stamboulis A, Baillie CA, Garkhail SK, Van Melick HGH, PeijsT. Environmental durability of flax fibres and their compositesbased on polypropylene matrix. Appl Comp Mater2000;7:273-94.[8] Zafeiropoulos NE, Williams DR, Baillie CA, Matthews FL.Engineering and characterisation of the interface in flax fibre/polypropylene composite materials. Part I. Development andinvestigation of surface treatments. Composites Part A2002;33: 1083-93.

[9] Avella M, Bozzi C, dell' Erba R, Focher B, Marzetti A,Martuscelli E. Steam-exploded wheat straw fibres as reinforcingmaterial for polypropylene-based composites. Angew MakromolChern 1995;233:149-66.

[10] Sanadi AR, Caulfield DF, Jacobson RE, Rowell RM. Renewableagricultural fibres as reinforcing fillers in plastics: mechanicalproperties of kenaf fibre-polypropylene composites. Ind EngChern Res 1995;34:1889-96.

[11] Felix JM, Gatenholm P. The nature of adhesion in composites ofmodified cellulose fibers and polypropylene. J Appl Polym Sci1991;42:609-20.

[12] Chuai C, Almdal K, Poulsen L, Plackett D. Conifer fibers asreinforcing materials for polypropylene-based composites. J ApplPolym Sci 2001;80:2833-41.

[13] Olsen DJ. Effectiveness of maleated polypropylenes as couplingagents for wood flour/polypropylene composites. Antec1991:1886-91.

[14] Arbelaiz A, Cantero G, Fernandez B, Garian P, Kenny JM,Mondragon I. Flax fiber surface modifications. effect on fibrephysico mechanical and flax/polypropylene interface properties.Polym Comp [in press].

[15] Kazayawoko M, Balatinecz JJ, Woodhams RT, Law S. Effectof ester linkages on the mechanical properties of wood fiber-polypropylene composites. J Reinf Plast Comp 1997;16:1383-406.

[16] Van De Velde K, Kiekens P. Influence of fibre surface charac-teristics on the flax/polypropylene interface. J Thermoplast CompMater 2001;14:244-60.

[17] Marcovich NE, Reboredo MM, Aranguren MI. Dependence ofthe mechanical properties of woodflour-polymer composites onthe moisture content. J Appl Polym Sci 1998;68:2069-76.

[18] Hornsby PR, Hinrichsen E, Tarverdi K. Preparation and prop-erties of polypropylene composites reinforced with wheat and flaxstraw fibres. Part II. Analysis of composite microstructure andmechanical properties. J Mater Sci 1997;32:1009-15.

[19] Kokta BV, Raj RG, Daneault C. Use of woodflour as filler inpolypropylene: studies on mechanical properties. Polym-PlastTech Eng 1989;28:247-59.

[20] Joseph K, Thomas S, Pavithran C. Effect of chemical treatmenton the tensile properties of short sisal fibre reinforced polyethyl-ene composites. Polymer 1996;37:5139-49.


11/11


[21] Sydenstricker THD, Mochnaz S, Amico SC. Pull-out and otherevaluations in sisal-reinforced polyester biocomposites. PolymTest 2003;22:375-80.

[22] Rana AK, Mandal A, Mitra BC, Jacobson R, Rowell R, BanerjeeAN. Short jute fiber-reinforced polypropylene composites: effectof compatibilizer. J Appl Polym Sci 1998;69:329-38.[23] Matias MC, De la Orden MU, Gonzalez Sanchez C, MartinezUrreaga J. Comparative spectroscopic study of the modificationof cellulosic materials with different coupling agents. J ApplPolym Sci 2000;75:256-66.

[24] Cantero G, Arbelaiz A, Llano-Ponte R, Mondragon 1. Effects offibre treatment on wettability and mechanical behaviour of flax/polypropylene composites. Comp Sci Tech 2003;63:1247-54.

[25] Eastman Chemical Company data sheet. Publication APG-l,September 1997.

[26] Thwe MM, Liao K. Durability of bamboo-glass fiber reinforcedpolymer matrix hybrid composites. Comp Sci Tech2003;63:375-87.

[27] Rozman HD, Peng GB, Mohd Ishak ZA. The effect ofcompounding techniques on the mechanical properties of oilpalm empty fruit bunch-polypropylene composites. J Appl PolymSci 1998;70:2647-55.

[28] Oksman K, Clemons C. Mechanical properties and morphologyof impact modified polypropylene-woodflour composites. J ApplPolym Sci 1998;67:1503-13.

[29] Arbelaiz A, Fernandez B, Cantero G, Llano-Ponte R, Valea A,Mondragon 1. Mechanical properties of flax fiber/polypropylenecomposites: influence of fibre/matrix modification and glass fibrehybridization. Composites Part A (sent for publishing).

[30] Felix JM, Gatenholm P, Schreiber HP. Controlled interactions incellulose-polymer composites. I: Effect on mechanical properties.Polym Comp 1993;14:449-57.

[31] Gassan J, Bledzki AK. Possibilities to improve the properties ofnatural fibre reinforced plastics by fibre modification-jute poly-propylene composites. Appl Comp Mater 2000;7:373-85.

[32] Qiu W, Zhang F, Endo T, Hirotsu T. Preparation and charac-teristics of composites of high-crystalline cellulose with polypro-pylene: effects of maleated polypropylene and cellulose content. JAppl Polym Sci 2003;87:337-45.

[33] Keener TJ, Stuart RK, Brown TK. Maleated coupling agentsfor natural fibre composites. Composites Part A2004;35:357-62.

[34] Joly C, Kofman M, Gauthier R. Polypropylene/cellulosic fibercomposites: chemical treatment of the cellulose assuming com-patibilization between the two materials. JMS-Pure Appl Chern1996;A33: 1981-96.

[35] Van den Oever MJA, Bos HI.. Critical fibre length and apparentinterfacial shear strength of single flax fibre polypropylenecomposites. Adv Compos Lett 1998;7:81-5.

[36] Gassan J, Bledzki AK. The influence of fibre-surface treatment onthe mechanical properties of jute-polypropylene composites.Composites Part A 1997;28A:lOOl-5.

[37] Van den Oever MJA, Bos HL, Van Kemenade MJJM. Influenceof the physical structure of flax fibres on the mechanicalproperties of flax fibre reinforced polypropylene composites.Appl Comp Mater 2000;7:387-402.

[38] Rezadoust AM, Esfandeh M. Evaluation of two-roll mill methodfor preparing short glass fibre reinforced NBR-phenolic compos-ites. Polym Polym Compos 2001;9:403-7.

[39] Ali R, Iannace S, Nicolais L. Effect of processing conditions onmechanical and viscoelastic properties of biocomposites. J ApplPolym Sci 2003;88:1637-42.

[40] Baiardo M, Zini E, Scandola M. Flax fiber-polyester composites.Composites Part A 2004;35:703-10.

[41] Joseph PV, Oommen Z, Joseph K, Thomas S. Melt rheologicalbehaviour of short sisal fibre reinforced polypropylene compos-ites. J Thermoplast Comp Mater 2002;15:89-114.

[42] Gauthier R, Joly C, Coupas AC, Gauthier H, Escoubes M.Interfaces in polyolefin/cellulosic fiber composites: chemicalcoupling, morphology, correlation with adhesion and aging inmoisture. Polym Comp 1998;19:287-300.

[43] Harikumar KR, Joseph K, Thomas S. Jute sack cloth reinforcedpolypropylene composites: mechanical and sorption studies. JReinf Plast Comp 1999;18:346-72.

[44] George J, Bhagawan SS, Thomas S. Effects of environment on theproperties of low-density polyethylene composites reinforced withpineapple-leaf fibre. Comp Sci Tech 1998;58:1471-85.

[45] Tajvidi M, Ebrahimi G. Water uptake and mechanical charac-teristics of natural filler-polypropylene composites. J Appl PolymSci 2003;88:941-6.[46] Marcovich NE, Reboredo MM, Aranguren M1. Moisture diffusionin polyester-woodflour composites. Polymer 1999;40:7313-20.

[47] Klason C, Kubat J, Stromvall HE. The efficiency of cellulosicfillers in common thermoplastics. Part 1. Filling without process-ing aids or coupling agents. Int J Polym Mater 1984;10:159-87.

[48] Imaz JJ, Rodriguez JL, Rubio A, Mondragon 1. Hydrothermalenvironment influence on water diffusion and mechanical behav-iour of carbon fibre/epoxy laminates. J Mater Sci Lett1991;10:662-5.

[49] Cabanelas JC, Prolongo SG, Serrano B, Bravo J, Baselga J. Waterabsorption in polyaminosiloxane-epoxy thermosetting polymers.J Mater Process Technol 2003;143-144:311-5.

[50] Sreekala MS, Kumaran MG, Thomas S. Water sorption in oilpalm fiber reinforced phenol formaldehyde composites. Compos-ites Part A 2002;33:763-77.

[51] Joseph PV, Rabello MS, Mattoso LHC, Joseph K, Thomas S.Environmental effects on the degradation behaviour of sisal fibrereinforced polypropylene composites. Comp Sci Tech2002;62: 1357-72.

[52] Karad SK, Jones FR, Attwood D. Moisture absorption bycyanate ester modified epoxy resin matrices. Part 1. Effect ofspiking parameters. Polymer 2002;43:5209-18.

[53] Espert A, Vilaplana F, Karlsson S. Comparison of waterabsorption in natural cellulosic fibres from wood and one-yearcrops in polypropylene composites and its influence on theirmechanical properties. Composites Part A 2004;35:1267-76.

[54] Caulfield DF, Feng D, Prabawa S, Young RA, Sanadi AR.Interphase effects on the mechanical and physical aspects ofnatural fibre composites. Angew Makromol Chern1999;272:57-64.

[55] Singh B, Gupta M, Verma A. The durability of jute fibre-reinforced phenolic composites. Comp Sci Tech 2000;60:581-9.

[56] Sreekala MS, Thomas S. Effect of fibre surface modification onwater-sorption characteristics of oil palm fibres. Comp Sci Tech2003;63:861-9.

[57] Joseph K, Thomas S, Pavithran C, Brahmakumar M. Tensileproperties of short sisal fibre-reinforced polyethylene composites.J Appl Polym Sci 1993;47:1731-9.

Pdn.sciencedirect.com Science Ob

Documents

Transcript of Pdn.sciencedirect.com Science Ob