IndustrialCropsandProducts84(2016)230–240
Contents lists available atScienceDirect
Industrial Crops and Products
j o u r n a l h o m e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / i n d c r o p
Modeling and analysis of the creep behavior of jute/green epoxy composites incorporated with chemically treated pulverized nano/micro jute fibers
Abdul Jabbar
a, Jiˇrí Militk ´y
a,∗, Bandu Madhukar Kale
a, Samson Rwawiire
a, Yasir Nawab
b, Vijay Baheti
aaDepartmentofMaterialEngineering,TechnicalUniversityofLiberec,Studentská2,46117Liberec,CzechRepublic
bDepartmentofFabricManufacturing,NationalTextileUniversity,Faisalabad37610Pakistan
a r t i c l e i n f o
Articlehistory:
Received7November2015
Receivedinrevisedform7December2015 Accepted19December2015
Keywords:
Jutefiber
Polymercomposites Creep
Greenepoxy Compressionmolding Pulverization
a b s t r a c t
Thispaperreportsthecreepbehaviorofalkalitreatedjute/greenepoxycompositesincorporatedwith variousloadings(1,5and10wt%)ofchemicallytreatedpulverizedjutefibers(PJF)atdifferentenvi- ronmenttemperatures.Compositeswerepreparedbyhandlayupmethodandcompressionmolding technique.Thecreepanddynamicmechanicaltestswereperformedinthree-pointbendingmodeby dynamicmechanicalanalyzer(DMA).TheincorporationofPJFisfoundtosignificantlyimprovethecreep resistanceandstrainrateofcomposites.Threecreepmodelsi.e.Burger’smodel,Findley’spowerlaw modelandasimplertwo-parameterpowerlawmodelwereusedtomodelthecreepbehaviorinthis study.Thetimetemperaturesuperpositionprinciple(TTSP)wasappliedtopredictthelong-termcreep performance.TheFindley’spowerlawmodelwasfoundtobesatisfactoryinpredictingthelong-term creepbehavior.Dynamicmechanicalthermalanalysis(DMTA)resultsrevealedtheincreaseinstorage modulus,glasstransitiontemperatureandreductioninthetangentdeltapeakheightofcompositeswith higherloadingofPJF.
©2015ElsevierB.V.Allrightsreserved.
1. Introduction
Naturalfiberpolymercomposites(NFPC)areincreasinglyused nowadaysinindustrialapplicationsasasubstituteofpolymercom- positesmadewithmostlyusedsyntheticfiberssuchasglassand carbonetc.duetotheirenvironmentalandeconomicbenefits.Nat- uralfibersarerenewable,biodegradable,costeffective,safetouse, availableinhugequantities,lowfossil-fuelenergyrequirements andthemostimportantlytheirhighspecificstrength toweight ratio(JohnandAnandjiwala,2009;Mishraetal.,2003).Thisisof distinctiveimportanceespeciallyininteriortransportationappli- cationsasitleadstoreductionofvehicleweightforhigherfuel efficiencyandenergy saving.Polymercompositesusedinengi- neeringapplicationsareoftensubjectedtostressforalongtime andathightemperatures.Creep(aprogressivedeformationofa materialatconstantstress)isveryimportantend-usepropertyfor materialapplicationsrequiringlongtermdurabilityandreliabil- ity(KremplandKhan,2003).Considerablestudiescanbefoundin
∗ Correspondingauthor.
E-mailaddress:jiri.militky@tul.cz(J.Militk ´y).
literatureonthecreepbehaviorofnaturalfiberpolymercompos- ites.Differentmathematicalmodelingtechniqueshavealsobeen appliedtoanalyzethecreepbehaviorofcompositematerials(Acha etal.,2007;BledzkiandFaruk,2004;Haoetal.,2014;Jiaetal., 2011;MarcovichandVillar,2003;Xuetal.,2010).Researchershave alsotriedtostudythecreepbehaviorofpolymercompositesby additionofdifferentkinds offillersinmatrices(Jiaetal.,2011;
Shenetal.,2004;Siengchin,2013;SiengchinandKarger-Kocsis, 2009;Yangetal.,2006a,b).Additionofnano/microfillerstopoly- mershasshownimprovementsinthestrengthandstiffnessofthe resultingcomposites,however,researchshowsthatthesefillers tendtoplasticizethecomposites.Totheauthor’sbestknowledge, thereisnostudyavailableinopenliteratureonthecreepbehavior ofalkalitreatedjutereinforcedgreenepoxycompositesincorpo- ratedwithchemicallytreatedpulverizedjutefibersasreinforcing fillers.Amongallthenaturalreinforcingmaterials,juteappearsto beapromisingfiberduetoitshightoughnessandaspectratioin comparisonwithothernaturalreinforcements(Achaetal.,2007) andoccupiesthesecondplaceintermsofworldproductionlevels ofcellulosicfibers(Caietal.,2000).Theobjectiveofthepresent studyistoinvestigatetheincorporationofpulverizednano/micro jutefiberspreparedfromwastejuteonthecreepbehaviorunder
http://dx.doi.org/10.1016/j.indcrop.2015.12.052 0926-6690/©2015ElsevierB.V.Allrightsreserved.
theconditionsofdifferenttemperaturesanddynamic mechani- calbehaviorofalkalitreatedwovenjute/greenepoxycomposites.
Themodelingofexperimentalcreepdataissatisfactorilyconducted usingdifferentcreepmodels.Furthermore,thetime-temperature superpositionprinciple(TTSP)isemployedinordertopredictthe long-termcreepbehavior ofcompositesbased onexperimental data.
2. Materialsandmethods
2.1. Materials
Jutewovenfabricproduced fromtossa jute(Corchorusolito- rius)fibershavinganarealdensityof600gm−2with5-endsatin weavedesignwasproducedonashuttleloom.Warpandweftden- sitiesofthefabricwere6.3threadspercmand7.9threadspercm respectively.Wastejutefibers,sourcedfromajutemill,wereused forpulverization.GreenepoxyresinCHS-EpoxyG520andhard- enerTELALIT0600weresuppliedbySpolchemie,CzechRepublic.
Sodiumhydroxide(NaOH)wassuppliedbyLach-Ner,CzechRepub- lic.Sodiumsulfate(Na2SO4)andsodiumhypochlorite(NaOCl)were suppliedbySigma–Aldrich,CzechRepublic.
2.2. Chemicalpretreatment
Theuntreated fabric wasfirst washed with2wt% non-ionic detergentsolutionat70◦Cfor1.0hpriortosurfacetreatmentsto removeanydirtandimpuritiesstickingtothesurface.Thejutefab- ricandwastejutefiberswerethenimmersedseparatelyin2%NaOH solutionfor1hat80◦Cmaintainingaliquorratioof15:1.Alkali treatedwastejutefiberswerefurthertreatedwith7g/lNaOClsolu- tionatroomtemperaturefor2hunderpH10–11andsubsequently antichloredwith0.1%Na2SO4at50◦Cfor20min.Bothfabricand wastefibers,afterchemicalpretreatment,werethenwashedwith freshwaterseveraltimesuntilthefinalpHwasmaintainedat7.0 andthenallowedtodryatroomtemperaturefor48handat100◦C inanovenfor2h.
2.3. Pulverizationofjutefibers
Pulverizationofchemicallytreatedwastejutefiberswascarried outusingahigh-energyplanetaryballmillofFritschpulverisette7.
Pulverizationprocessreliesontheprincipleofenergyreleaseatthe pointofimpactbetweenballsaswellasonthehighgrindingaction createdbyfrictionofballsonthewall(BahetiandMilitky,2013).
Thesinteredcorundumcontainerof80mlcapacityandzirconium ballsof10mmdiameterwerechosenfor1hofpulverization.The balltomaterialratio(BMR)waskeptat10:1andthespeedwas keptat850rpm.
2.4. Preparationofcomposites
Thecompositeswerepreparedbyhandlayupmethod.Theresin andhardenerweremixedinaratioof100:32(byweight)according tomanufacturerrecommendationsandthenweighedamountsof pulverizedjutefibers(PJF)under1,5and10wt%weremechanically mixedwithepoxyresinatroomtemperatureuntilahomogeneous mixturewasobtained.Thepreparedresin/PJFmixturewaspoured onfabriclayersandspreadoutbyahandroller.Thegentlerolling actionofhandrollerconfirmedthewettingofjute fabrics.Each compositelaminatecomprised3layersofjutefabricwithorienta- tionofeachlayerinthesamedirection.Thecompositelayupalong withTeflonsheetsweresandwichedbetweenapairofsteelplates andcuredat120◦Cfor1.0hinamechanicalconvectionovenwith predeterminedweightonittomaintainuniformpressureofabout 50kPa(Mishraetal.,2014).Thefibervolumefraction(Vf)ofall
compositeswasintherangeof0.25–0.27.Thepreparedcomposite samplesweredesignatedasU(untreated),A-0%(alkalitreatedjute fabricwith0wt%ofPJF),1%,5%and10%(alkalitreatedjutefabric with1,5,10wt%ofPJF)respectively.
2.5. Characterizationandtesting
2.5.1. Characterizationofpulverizedjutefibers
Theuntreatedandtreatedjutefabricandfiberswereanalyzed byFTIRspectroscopy.AThermoFisherFTIRspectrometer,model NicoletiN10,wasusedinthisstudy.Thespectrometerwasusedin theabsorptionmodewitharesolutionof4cm−1.Particlesizedistri- butionofpulverizedjuteparticleswasstudiedonMalvernZetasizer nanoseriesbasedondynamiclightscatteringprincipleofBrown- ianmotionofparticles.Deionizedwaterwasusedasdispersion mediumanditwasultrasonicatedfor5minwithBANDELINultra- sonicprobeSONOPLUSbeforecharacterization.Refractiveindexof 1.52wasusedtocalculateparticlesizeofpulverizedjute.Inaddi- tion,morphologiesofpulverizedjutefiberswereobservedwith Vega-TescanTS5130ScanningElectronMicroscopeat30KVaccel- eratingvoltage.ThesurfaceoffiberswasgoldcoatedpriortoSEM inspection.
2.5.2. Characterizationofcomposites
Short-termcreeptestswereperformedinthreepointbending modeattemperatures40◦C,70◦Cand100◦CusingQ800Dynamic mechanicalthermalanalysis(DMTA)instrumentofTAinstruments (NewCastleDL,USA)for30min.Thestaticstressof2.0MPawas appliedat thecenter point of long sideof the samplethrough thesamplethicknessfor30minafterequilibratingatthedesired temperatureandcreepstrainwasmeasuredasafunctionoftime.
Thestaticstresswasselectedafterperformingastrainsweeptest, wherethelinearviscoelasticregionwasdefinedforeach ofthe compositesensuringthatthecreeptestswereconductedinthe linearviscoelasticregion.TheTTSP wasselectedfor short-term creeptestsperformedatvarioustemperaturesforjute/greenepoxy compositesincorporatedwithvariouscontentsofPJF. Thetem- peraturerange was40–100◦C,in5◦Csteps,andtheisothermal testswererunonthesamespecimeninthetemperaturerange.
The2.0MPastresswasappliedfor10minateachtemperature.In everymeasurement,thespecimenwasequilibratedfor5minat eachtemperature,inordertoevenlyadjustforthecorrecttem- peratureofthespecimen.Thedynamicmechanicalpropertiesof composites weremeasuredin 3-pointbendingmodeusing the abovesameinstrument.Thetestingconditionswerecontrolledin thetemperaturerangeof35–200◦C,withaheatingrateof3◦C/min, fixedfrequencyof1Hz,preloadof0.1N,amplitudeof20m,and forcetrackof125%.Thesampleshavingathicknessof4.5–5mm, widthof12mmandspanlengthof50mmwereusedforbothcreep andDMTAtesting.Tworeplicatesamplesweretestedforeachtest conditionandaveragevaluesarereported.
2.6. Creepmodeling
Thenon-linearcurvefitfunctionoftheOriginPro9.0software wasusedformodelingthecreepcurvesandfittingthemodelsto theexperimentaldata.Theminimumsumofsquareddeviationof experimentaldatafromthecreepmodelsandcorrelationcoeffi- cient(R2)wereselectedascriterion(MelounandMilitky,2011).The correlationcoefficientvalueR2isdefinedasmodelsumofsquares dividedbytotalsumofsquares.Abettergoodness-to-fitisobtained whenR2iscloserto1.
232 A.Jabbaretal./IndustrialCropsandProducts84(2016)230–240
Fig.1. FTIRspectraofjutefibers.
3. Resultsanddiscussion
3.1. Characterizationofjutefibers
FTIRanalysiswascarriedouttoconfirmsomeremovalofnon- cellulosiccontents(e.g.hemicellulosesandlignin)fromthesurface ofjutefibersafterpretreatments.TheFTIRspectraofuntreatedand treatedjute areshownin Fig.1.Themajordifferenceobserved betweenthespectrais thedisappearance/reductionof peaksat
∼1730cm−1 and∼1240cm−1.Thepeakat∼1730cm−1 isdueto stretchvibrationofC Obondsincarboxylicacidandestercom- ponentsofcelluloseandhemicellulose(Morshedetal.,2010)and alsoachemicalgroupoflignin(Haqueetal.,2009;Tserkietal., 2005). The peak at ∼1240cm−1 is due to C O C asymmetric stretchingoftheacetylgroupoflignin(Liuetal.,2004).Thereduc- tion/disappearanceofthesepeaksconfirmthepartialremovalof hemicelluloseandopeningupoftheligninstructureinthejute fibersafterpretreatments.
Thejutefiberswerepulverizedtoparticleswithanaveragesize of1480nminwiderparticlesizedistributionasshowninFig.2a.
Fig.2bshowstheshapeandsurfacemorphologyofpulverizedjute fibers.ThePJFcanbeseeninirregularshapeandsizewithcertain aspectratioaswell.
3.2. Creepbehavior
3.2.1. Creepmodels
Fourparameters(ortheBurger’s)modelisoneofthemostly usedphysicalmodelstogivetherelationshipbetweenthemor- phologyofpolymercompositesandtheircreepbehavior(Findley et al.,1989; Ward and Sweeney, 2012). It is based ona series combinationofaMaxwellelementwithaKelvin–Voigtelement.
Thetotalcreepstrainisdividedintothreeseparateparts:εMthe instantaneouselasticdeformation(Maxwellspring),εKviscoelas- ticdeformation(Kelvinunit)andε∞viscousdeformation(Maxwell dash-pot).Thus,totalstrainasafunctionoftimecanberepresented bythefollowingequations:
ε (t)=εM+εK+ε∞ (1)
ε (t)= 0 EM+0
EK(1−e−EKt
⁄
K)+ 0Mt (2)
Whereε(t)isthecreepstrain,0isthestress,tisthetime,EMand EKaretheelasticmoduliofMaxwellandKelvinsprings,andM andKaretheviscositiesofMaxwellandKelvindashpots.K/EK
isusuallydenotedas,theretardationtimerequiredtogenerate 63.2%deformationintheKelvinunit(Yangetal.,2006b).εMisa constantvalueanddoesnotchangewithtime.εKrepresentsthe earlieststageofcreepandattainsasaturationvalueinshorttime andε∞representsthetrendinthecreepstrainatsufficientlylong time,andappearssimilartothedeformationofaviscousliquid obeyingNewton’slawofviscosity.ThefourparametersEM,EK,M,
KcanbeobtainedbyfittingtheEq.(2)totheexperimentaldata andcanbeusedtodescribethecreepbehaviorofcomposites.The creeprateofviscoelasticmaterialscanbeobtainedbytakingthe derivativeofEq.(2).
dε (t) dt =0
EK(e−EKt
⁄
K)+ 0M (3)
TheFindley’spowerlawmodelisanempiricalmathematical modelusedtosimulatethecreepbehaviorofpolymercomposites.
ThemodelcanberepresentedbythefollowingEq.(4)(Findleyetal., 1989);
ε (t)=atb+ε0 (4)
Where,aandbarethematerialconstantsandε0istheinstanta- neousstrain.TheabilityofFindley’spowerlawmodeltosimulate thecreepdatahasbeenfoundtobesatisfactoryinseveralstud- ies(Jiaetal.,2011;PlaseiedandFatemi,2008;Yangetal.,2006b).
However,thismodelisnotabletoexplainthecreepmechanism ofmaterial.Atwoparameterempiricalpowerlawmodelhasalso beenusedinsomestudies(Tajvidietal.,2005;Xuetal.,2010)to simulatethecreepdate.Ithastheform;
ε (t)=atb (5)
Where,aandbarethematerialconstantsandε0istheinstanta- neousstrain.
Thelongtermcreepisanimportantparametertoevaluatethe end-useperformanceofnaturalfiberpolymercompositesbutit
Fig.2. (a)Particlesizedistributionand(b)SEMimageofjutefibersafterpulverization.
isoftenimpracticaltoperformacreeptestforanextremelylong period oftime. Time-temperaturesuperposition (TTS)isone of thecommonestimationtechniquestopredictthelongtermcreep behaviorbyshiftingthecurves fromtestsatdifferenttempera- tureshorizontallyalongthelogarithmictimeaxistogeneratea singlecurveknownasmastercurve(WardandSweeney,2012).
Theshiftingdistanceiscalledshiftfactor.
3.2.2. Shorttermcreep
Fig.3showsthecreepstrainsforjutecompositesasafunction oftimewith0,1,5,10wt%ofPJFcontentatthreedifferenttem- peratureconditions.Itisvisiblyapparentthatthecompositeshave lowinstantaneousdeformationεMandcreepstrainat40◦Cdue tohigherstiffnessofcompositesbutthisdeformationincreasesat highertemperaturesduetodecreaseincompositesstiffness.The creepstrainofallcompositesalsoincreasedathighertempera- turesbuttheuntreatedjutecompositeswasaffectedmorethan
theothers. The creepstrain of alkalitreated with0% PJF com- positeislessthanuntreatedone.Thismaybeexplaineddue to increaseinsurfaceroughnessofjutefabricafteralkalitreatment and decrease infrictional slippageof matrixpolymer chainsat thefiber/matrixinterfaceresultinginlesscreepdeformationthan untreatedcomposite.Theleastcreepstrainisshownbycompos- iteincorporatedwith10%PJFatalltemperaturesfollowedby5%
and 1%PJF incorporatedcomposites. At100◦C,5% and 10%PJF compositeshavealmostsameinstantaneouselasticdeformation but10%composite haslessviscousdeformationovertime. This maybeattributedtogreaterinhibitionofslippageandreorienta- tionofpolymerchainwithincreasingcontentsofPJF.TheBurger’s modelcurvesshowasatisfactoryagreementwiththeexperimen- taldata(Fig.3).ThefourparametersEM,EK,M,K of Burger’s model,usedtofittheEq.(2)totheexperimentaldata,aresum- marizedinTable1.Allfourparameterswerefoundtodecreasefor allcompositesastemperatureincreased(Table1).EMcorresponds
234 A.Jabbaretal./IndustrialCropsandProducts84(2016)230–240
Fig.3.CreepcurvesofcompositesincorporatedwithdifferentloadingsofPJFatdifferenttemperatures.
Fig.3.Continued.
Table1
SimulatedfourparametersinBurger’smodelforshorttermcreepofthecomposites.
Temperature Parameters Compositetypes
Untreated Alkali-0% 1% 5% 10%
40◦C EM(MPa) 2477.24±78.3 3259.41±138.0 3492.53±133.8 3810.53±133.6 3774.75±149.1 EK(MPa) 23876.27±10854.5 38244.62±19726.9 42496.38±21110.6 44220.12±21281.1 40549.80±19016.6
M(Pas) 2.72E7±1.33E7 3.54E7±1.22E7 4.77E7±2.04E7 4.54E7±1.97E7 5.11E7±2.47E7
K(Pas) 2.11E6±2.38E6 1.37E6±2.09E6 1.73E6±2.5E6 2.53E6±3.49E6 1.90E6±2.62E6
SS* 2.65033E-9 1.29037E-9 1.04302E-9 9.92292E-10 1.08146E-9
Adj.R2 0.97524 0.97061 0.96424 0.96929 0.96304
70◦C EM(MPa) 1985.89±91.9 2403.61±102.7 2921.19±118.5 3116.67±118.6 3323.10±131.9 EK(MPa) 7790.84±2752.5 11972.90±4497.3 14228.47±5360.3 16946.57±6037.4 15615.45±5726.5
M(Pas) 9.48E6±3.81E6 1.32E7±5.14E6 1.98E7±9.71E6 2.33E7±1.08E7 2.27E7±1.13E7
K(Pas) 956179.43±6.90E5 1.32E6±1.09E6 1.71E6±1.34E6 1.81E6±1.45E6 1.94E6±1.44E6
SS* 1.08937E-8 5.93037E-9 3.81543E-9 2.74811E-9 2.87718E-9
Adj.R2 0.98881 0.98689 0.9851 0.9847 0.98596
100◦C EM(MPa) 1380.05±294.3 1743.43±306.1 2417.83±218.3 2653.56±183.6 2616.45±123.7 EK(MPa) 665.35±117.7 865.63±121.6 3543.54±958.0 6087.72±1783.7 8457.41±2683.5
M(Pas) −4.13E20±0.0 −6.00E32±0.0 7.44E6±3.92E6 1.08E7±5.26E6 1.31E7±5.97E6
K(Pas) 219089.75±1.03E5 255733.51±9.91E4 457813.67±2.43E5 705184.01±4.39E5 1.11E6±6.84E5
SS* 8.30033E-7 3.37586E-7 2.87859E-8 1.31777E-8 6.79893E-9
Adj.R2 0.97565 0.98358 0.99023 0.98826 0.98971
SS*:Sumofsquareddeviations.
totheelasticityofthecrystallizedzonesinasemicrystallizedpoly- mer.Comparedtotheamorphousregions,thecrystallizedzones aresubjectedtoimmediatestressduetotheirhigherstiffness.The instantaneouselasticmodulusisrecoveredimmediatelyoncethe stressis removed.EKisalsocoupledwiththestiffnessofmate- rial.ThedecreaseinparametersEM,EKresultedfromtheincrease intheinstantaneousandtheviscoelasticdeformationsastemper- atureincreased.TheviscosityMcorrespondstodamageinthe crystallizedzonesandirreversibledeformationintheamorphous regionsandtheviscosityKisassociatedwiththeviscosityofthe amorphousregionsinthesemicrystallizedpolymer(Militk ´yand Jabbar,2015).The decreaseinviscosityparametersM,Kpro- poseanimprovementinthemobilityofmolecularchainsathigher
temperature.Theparametersforuntreatedandalkalitreatedwith 0%PJFcompositeshaveundergonealargestdecrease,resultingin highercreepstrain.ThecompositesincorporatedwithPJF,espe- cially5and10%,havecomparativelybettervaluesofparameters particularlyMwhichisrelatedtothelongtermcreepstrainand validateslesstemperaturedependenceofthesecomposites(Fig.3).
TheviscosityMincreaseswiththeincreaseinPJF%andpermanent deformationdecreases.Fig.4aandbcomparesthecreepstrainand strainrateofuntreatedand10%PJFcompositesatvarioustem- peratures.Comparatively,temperaturehadmoreinfluenceonthe creepdeformationofuntreatedjutecompositethanthatof10%PJF composite.
236 A.Jabbaretal./IndustrialCropsandProducts84(2016)230–240
Fig.4.Creepstrain(a)andstrainrate(b)ofuntreatedand10%PJFcompositesatdifferenttemperatures.
3.2.3. Time-temperaturesuperposition(TTS)
Thecreepcurvescorrespondingtodifferenttemperaturelev- els were shifted along the logarithmic time axis according to time–temperaturesuperpositionprinciple usingTAInstruments ThermalAdvantageTMsoftwaretogenerateamastercurveataref- erencetemperatureof40◦C.Theshiftingprocedureofthiscurve
obeystheWilliams–Landel–Ferry(WLF)equation.TheWLFequa- tionisgivenbyEq.(6);
log˛T= −C1(T−T0)
C2+ (T−T0) (6)
Where␣T isthehorizontal(ortime)shiftfactor,C1 and C2 are constants,T0 isthe referencetemperature(K) andT is thetest
Fig.5.TTSmastercurvesforcreepofthecompositesincorporatedwithdifferentloadingsofPJFatareferencetemperatureof40◦C.
temperature(K).Themastercurves,whichgiveanindicationof long-termcreepperformanceofcomposites,areplottedinlog–log scaleandpresentedinFig.5.Themastercurvesshowbettercreep resistanceofcompositeswithincreasingcontentofPJF.Itisobvi- ousthatthebestlongtermperformanceisshownbycomposites incorporatedwith5and10%PJFindicatingtheirgoodreinforcing effectiveness.Itisalsointerestingtonotethatabovelog-time4.0s, thecreepdeformationofuntreated,0and1%compositesshowa fastertendencyofincreasecompared to5and10%PJFcompos- ites.Thesefindingsshowthatunderthesmallstress,thematerials enteredintoaviscoelasticstateoveranextremelylongperiodof time,andinviscoelasticstatetheroleof1%incorporationofPJFin thereinforcementeffectivenessislessthanthatof5and10%PJF.
Fig.6showsthatthepredictionabilityoftheBurger’smodel, Findley’smodelandtwo parameterpowerlawmodel(Eqs. (2), (4)and(5))forthelog-termcreepbehaviorofcompositesatthe referencetemperatureof40◦C.TheincorporationofPJFincom- positesdecreasedthelongtermcreepstraincompareduntreated and0%filledalkalitreatedcomposites.Itcanbeclearlyseenthat Findley’smodel is good to predictthe long-termcreep perfor- mance. However, this model provides the adequate prediction abilitywithinthesteady statecreepandgiven time intervalas revealed by some researchers (Siengchin, 2013; Siengchin and Karger-Kocsis,2009)whileforlongertimedurationthecalculated datamayshowconsiderabledeviationfromtheexperimentaldata.
ThepredictionoftheBurger’smodelshowssomedeviationand thatoftwoparameterpowerlawmodel,showsalittlelargedevi- ationfromtheexperimentaldata.Similarfindingswerereported byotherresearchers(Haoetal.,2014;Jiaetal.,2011).Thesim- ulatedparametersofBurger,Findleyandtwoparameterspower
lawmodelsaresummarizedinTable2.TheparametersofBurger model,resultedfromfittingcreepcurves,areverydifferentfrom thoseoftheshorttermcreeptests.It canbeseenfromTable2 thatalltheBurgermodelparametersincreasewiththeincrease inloadingofPJF%.TheM,whichdetermineslong-termcreep,is lowestforuntreatedcompositeandhighestfor10%PJFcomposite.
Therefore,theuntreatedcompositeshowsthehighestand10%PJF compositeshowsthelowestcreepdeformation.Itisalsoobvious forFindleymodelthattheparametera(reflectingshort-termcreep) increasedandparametersε0 (reflectingtheinstantaneousinitial creepstrain)andb(reflectinglog-termcreep)decreasedwiththe increasingcontentofPJFwhichindicatesanenhancedlong-term creepperformancewithPJF loading.Similarly, fortwo parame- terpowerlawmodel,parametera(reflectingshort-termcreep) increasedandparameterb(reflectinglog-termcreep)decreased withtheincreasingcontentofPJF.
3.3. Dynamicmechanicalthermalproperties
Dynamicmechanicalthermalanalysiscancharacterizethevis- coelasticpropertiesofthematerialsanddeterminetheinformation ofstoragemodulus,lossmodulus(theenergydissipationassoci- atedwiththemotionofpolymerchains)andlossfactor(tandelta) ofpolymer compositeswithinthemeasuredtemperaturerange (Wangetal.,2015).Thevariationofstoragemodulus(E)ofcom- positesincorporatedwithdifferentcontentofPJFasafunctionof temperatureatfrequencyof1HzisshowninFig.7.Itcanbeseen fromFig.7athatthereisagradualfallinthestoragemoduliwith temperature,whichshouldberelatedwithanenergydissipation phenomenoninvolvingcooperativemotionsofthepolymerchains
238 A.Jabbaretal./IndustrialCropsandProducts84(2016)230–240
Fig.6. Long-termcreeppredictionofcompositesby(a)Burger’smodel,(b)Findleymodeland(c)twoparameterpowerlawmodelat40◦C.
Table2
SimulatedparametersofBurger’smodel,Findley’spowerlawmodelandtwoparameterspowerlawmodelforlongtermcreeppredictionofthecompositesat40◦C.
Temperature Parameters Compositetypes
Untreated Alkali-0% 1% 5% 10%
Burger’s model
EM(MPa) 2713.76±193.6 3002.45±184.0 2904.39±179.0 3478.01±191.0 3517.07±125.5 EK(MPa) 4781.98±2005.8 5470.29±2194.5 6271.53±2337.0 15949.06±7113.4 31490.67±16174.2
M(Pas) 2.64E9±1.29E9 2.94E9±1.29E9 4.01E9±2.19E9 9.44E9±6.93E9 2.10E10±2.04E10
K(Pas) 1.0227E8±8.73E7 1.35E8±1.05E8 9.24E7±7.83E7 6.76E7±9.00E7 9.49E7±1.53E8
SS* 4.86566E-7 3.00571E-7 2.96188E-7 1.2249E-7 4.56987E-8
Adj.R2 0.9696 0.97551 0.9658 0.92279 0.88442
Findley’s model
a 1.05E-5±4.83E-06 6.49E-6±3.69E-06 1.79E-5±1.14E-05 3.49E-5±3.45E-05 3.35E-5±5.92E-04 b 0.34341±3.39E-02 0.3699±4.22E-02 0.28033±4.63E-02 0.17143±6.61E-02 0.1284±2.27E+00 ε0 6.70E-4±2.68E-05 6.17E-4±2.58E-05 6.13E-4±3.82E-05 5.00E-4±5.74E-05 5.11E-4±9.03E-03
SS* 3.83097E-8 4.27255E-8 4.37949E-8 1.91548E-8 1.30874E-8
Adj.R2 0.99761 0.99653 0.99496 0.98797 0.96701
Two parameter powerlaw model
a 4.63E-4±1.41E-04 4.24E-4±1.39E-04 4.82E-4±1.00E-04 4.85E-4±3.61E-05 5.22E-4±2.08E-05 b 0.08443±3.37E-02 0.08145±3.64E-02 0.06828±2.37E-02 0.03723±9.03E-03 0.01951±5.01E-03
SS* 2.4937E-6 2.31209E-6 1.0253E-6 9.22041E-8 2.50976E-8
Adj.R2 0.8452 0.8128 0.88238 0.94226 0.93694
SS*:sumofsquareddeviations.
withtemperature(Fengetal.,2011).Theincreaseinstoragemodu- lusoverthewholetemperaturerangewasobservedforcomposites incorporatedwithdifferentloadingsofPJF,forexample,additionof 1,5and10%PJFcausesasignificantincreaseof∼18%,22%and43%
inthestoragemodulusrespectivelyat35◦C.Moreover,thestorage moduluscurvesofcompositeshavebeenshiftedtohighertemper- aturesafteradditionofthePJF,particularly5and10%loading.This significantimprovementinstoragemodulusisduetobetterrein- forcingeffectofPJFleadingtoincreasedstiffnessandthemobility restrictionofpolymerchains(Jiaetal.,2011).Thechangeinloss
factor(tanı,theratiooflossmodulustocorrespondingstorage modulus)ofcompositeswithdifferentloadingofPJFasafunction oftemperatureisshowninFig.7c.Untreatedcompositedisplayed ahighertanıpeakvaluethanothers.Thismaybeattributedto moreenergydissipationduetofrictionaldampingattheweaker untreatedfiber/matrixinterface.Thetemperatureatwhichtan␦ attainsamaximumvaluecanbereferredtoastheglasstransition temperature(Tg)(ShanmugamandThiruchitrambalam,2013).A positiveshiftinTgcanbeobservedforallcompositesincorporated withPJFcomparedtountreatedcompositeThelowertan␦peak
Fig.7.DynamicmechanicalpropertiescompositesincorporatedwithdifferentloadingsofPJF;(a)storagemodulus,(b)lossmodulus,(c)tandelta.
Table3
TgvaluesobtainedfromEcurves.
Composites TgfromEmaxcurve(◦C)
Untreated 110.10
Alkali-0% 110.60
1% 114
5% 123
10% 137
heightisshownbycompositeincorporatedwith10%PJFfollowed by5%and1%PJFcomposites,exhibitingastrongfiber/matrixinter- facialinteractionswhichcanrestrictthesegmentalmovementof thepolymerchainsleadingtotheincreasedTg.
IthasbeenreportedthatTgvaluesobtainedfromlossmodulus (E)curvepeakaremorerealisticascomparedtothoseobtained fromlossfactor(tanı)(Akay,1993).ApositiveshiftinTgtohigher temperatureforallcompositesincorporatedwithPJFisobserved, i.e.Tgincreasedfrom110.1◦Cforuntreatedto∼110.6,114,123and 137◦Cforcompositesincorporatedwith0,1,5and10%PJFrespec- tivelyaspresentedinTable3andFig.7b.Thismaybeduetoreduced mobilityofmatrixpolymerchainsandbetterreinforcementeffect ofPJF.Ithasbeenreportedthatsystemscontainingmorerestric- tionsandahigherdegreeofreinforcementtendtoexhibithigher Tg(Almeidaetal.,2012).
4. Conclusion
Creepbehaviorofalkalitreatedwovenjute/greenepoxycom- positesincorporatedwithdifferentloadingsofchemicallytreated pulverizedjutefibers(PJF)waspresentedatvariousenvironment temperatures.Thecreepdeformationwasfoundtoincreasewith
temperature. The creep resistanceof composites was found to improvesignificantlywiththeincorporationofPJF.Themodeling ofcreepdatawassatisfactorilyconductedbyusingBurger’smodel, Findley’spowerlawmodelandasimplertwo-parameterpowerlaw model.TheBurger’smodelfittedwelltheshort-termcreepdata.
However,theFindley’spowerlawmodelwassatisfactoryforpre- dictingthelong-termcreepperformanceofcompositescompared toBurgerandtwoparametermodel.Themastercurves,generated byTTSprinciple,indicatedthepredictionofthelong-termperfor- manceofcomposites.Dynamicmechanicaltest resultsrevealed theincreaseinstoragemodulus,glasstransitiontemperatureand reduction in the tangent delta peak heightof PJF incorporated composites.Basedontheanalysisofresults,theimprovedcreep resistanceofthecompositeswerelikely attributedtotheinhib- itedmobilityofpolymermatrixmolecularchainsinitiatedbylarge interfacialcontactareaofPJFaswellastheirinterfacialinteraction withthepolymermatrix.
Acknowledgements
Thisworkwassupportedunderthestudentgrantscheme(SGS- 21085)byTechnicalUniversityofLiberec,CzechRepublic.Oneof theauthorsisthankful toDr.Ing.Antonín Potˇeˇsiland Ing.Petr HornikforprovidingDMAtestingfacilities.
References
Acha,B.A.,Reboredo,M.M.,Marcovich,N.E.,2007.Creepanddynamicmechanical behaviorofPP–jutecomposites:Effectoftheinterfacialadhesion.Compos.
PartAAppl.Sci.Manuf.38,1507–1516.
Akay,M.,1993.Aspectsofdynamicmechanicalanalysisinpolymericcomposites.
Compos.Sci.Technol.47,419–423.
240 A.Jabbaretal./IndustrialCropsandProducts84(2016)230–240
AlmeidaJr.,J.H.S.,OrnaghiJr.,H.L.,Amico,S.C.,Amado,F.D.R.,2012.Studyofhybrid intralaminatecuraua/glasscomposites.Mater.Des.42,111–117.
Baheti,V.,Militky,J.,2013.Reinforcementofwetmilledjutenano/microparticles inpolyvinylalcoholfilms.FibersPolym.14,133–137.
Bledzki,A.K.,Faruk,O.,2004.Creepandimpactpropertiesofwood fibre–polypropylenecomposites:influenceoftemperatureandmoisture content.Compos.Sci.Technol.64,693–700.
Cai,Y.,David,S.,Pailthorpe,M.,2000.Dyeingofjuteandjute/cottonblendfabrics with2:1pre-metalliseddyes.DyePigm.45,161–168.
Feng,Q.-P.,Shen,X.-J.,Yang,J.-P.,Fu,S.-Y.,Mai,Y.-W.,Friedrich,K.,2011.Synthesis ofepoxycompositeswithhighcarbonnanotubeloadingandeffectsoftubular andwavymorphologyoncompositestrengthandmodulus.Polymers52, 6037–6045.
Findley,W.N.,Davis,F.A.,Onaran,K.,1989.CreepandRelaxationofNonlinear ViscoelasticMaterials:WithanIntroductiontoLinearViscoelasticity.Dover Publications,Inc.,NewYork.
Hao,A.,Chen,Y.,Chen,J.Y.,2014.Creepandrecoverybehaviorof kenaf/polypropylenenonwovencomposites.J.Appl.Polym.Sci.131.
Haque,M.M.,Hasan,M.,Islam,M.S.,Ali,M.E.,2009.Physico-mechanicalproperties ofchemicallytreatedpalmandcoirfiberreinforcedpolypropylene
composites.Biores.Technol.100,4903–4906.
Jia,Y.,Peng,K.,Gong,X.-l.,Zhang,Z.,2011.Creepandrecoveryof
polypropylene/carbonnanotubecomposites.Int.J.Plast.27,1239–1251.
John,M.J.,Anandjiwala,R.D.,2009.Chemicalmodificationofflaxreinforced polypropylenecomposites.Compos.PartAAppl.Sci.Manuf.40,442–448.
Krempl,E.,Khan,F.,2003.Rate(time)-dependentdeformationbehavior:an overviewofsomepropertiesofmetalsandsolidpolymers.Int.J.Plast.19, 1069–1095.
Liu,W.,Mohanty,A.,Drzal,L.,Askel,P.,Misra,M.,2004.Effectsofalkalitreatment onthestructure,morphologyandthermalpropertiesofnativegrassfibersas reinforcementsforpolymermatrixcomposites.J.Mater.Sci.39,1051–1054.
Marcovich,N.E.,Villar,M.A.,2003.Thermalandmechanicalcharacterizationof linearlow-densitypolyethylene/woodflourcomposites.J.Appl.Polym.Sci.90, 2775–2784.
Meloun,M.,Militky,J.,2011.StatisticalDataAnalysis:APracticalGuide.
WoodheadPublishing,Limited.
Militk ´y,J.,Jabbar,A.,2015.Comparativeevaluationoffibertreatmentsonthecreep behaviorofjute/greenepoxycomposites.Compos.PartBEng.80,361–368.
Mishra,S.,Mohanty,A.K.,Drzal,L.,Misra,M.,Parija,S.,Nayak,S.K.,Tripathy,S.S., 2003.Studiesonmechanicalperformanceofbiofibre/glassreinforced polyesterhybridcomposites.Compos.Sci.Technol.63,1377–1385.
Mishra,R.,Baheti,V.,Behera,B.,Militky,J.,2014.Noveltiesof3-Dwoven compositesandnanocomposites.J.Text.Inst.105,84–92.
Morshed,M.,Alam,M.,Daniels,S.,2010.Plasmatreatmentofnaturaljutefibreby RIE80plusplasmatool.Plas.Sci.Technol.12,325.
Plaseied,A.,Fatemi,A.,2008.Tensilecreepanddeformationmodelingofvinyl esterpolymeranditsnanocomposite.J.Reinf.Plast.Compos.28,1775–1788.
Shanmugam,D.,Thiruchitrambalam,M.,2013.Staticanddynamicmechanical propertiesofalkalitreatedunidirectionalcontinuouspalmyrapalmleafstalk fiber/jutefiberreinforcedhybridpolyestercomposites.Mater.Des.50, 533–542.
Shen,L.,Phang,I.Y.,Chen,L.,Liu,T.,Zeng,K.,2004.Nanoindentationand morphologicalstudiesonnylon66nanocomposites.I.Effectofclayloading.
Polymers45,3341–3349.
Siengchin,S.,2013.Dynamicmechanicandcreepbehaviorsof polyoxymethylene/boehmitealuminananocompositesproducedby water-mediatedcompoundingEffectofparticlesize.J.Thermo.Compos.
Mater.26,863–877.
Siengchin,S.,Karger-Kocsis,J.,2009.Structureandcreepresponseoftoughened andnanoreinforcedpolyamidesproducedviathelatexroute:effectof nanofillertype.Compos.Sci.Technol.69,677–683.
Tajvidi,M.,Falk,R.H.,Hermanson,J.C.,2005.Time–temperaturesuperposition principleappliedtoakenaf-fiber/high-densitypolyethylenecomposite.J.
Appl.Polm.Sci.97,1995–2004.
Tserki,V.,Zafeiropoulos,N.,Simon,F.,Panayiotou,C.,2005.Astudyoftheeffectof acetylationandpropionylationsurfacetreatmentsonnaturalfibres.Compos.
PartAAppl.Sci.Manuf.36,1110–1118.
Wang,X.,Gong,L.-X.,Tang,L.-C.,Peng,K.,Pei,Y.-B.,Zhao,L.,Wu,L.-B.,Jiang,J.-X., 2015.Temperaturedependenceofcreepandrecoverybehaviorsofpolymer compositesfilledwithchemicallyreducedgrapheneoxide.Compos.PartA Appl.Sci.Manuf.69,288–298.
Ward,I.M.,Sweeney,J.,2012.MechanicalPropertiesofSolidPolymers.JohnWiley
&Sons.
Xu,Y.,Wu,Q.,Lei,Y.,Yao,F.,2010.Creepbehaviorofbagassefiberreinforced polymercomposites.Biores.Technol.101,3280–3286.
Yang,J.-L.,Zhang,Z.,Schlarb,A.K.,Friedrich,K.,2006a.Onthecharacterizationof tensilecreepresistanceofpolyamide66nanocomposites.PartI.Experimental resultsandgeneraldiscussions.Polymers47,2791–2801.
Yang,J.-L.,Zhang,Z.,Schlarb,A.K.,Friedrich,K.,2006b.Onthecharacterizationof tensilecreepresistanceofpolyamide66nanocomposites.PartII:modeling andpredictionoflong-termperformance.Polymers47,6745–6758.