AbstractComposite materials and layered structures based on nat-

ural plant fibers are increasingly regarded as an alternative toglass fiber reinforced parts. One of their major field of appli-cation can be found in structural components for theautomotive industry. Product examples are door trimpanels or instrument panels. For such applications anutmost impact strength is required in order to imple-ment a maximum of passenger safety by a good crashbehavior. The paper describes the effects of severalmaterial parameters such as fiber fineness or fleececomposition as well as the impact of the process condi-tions on this important composite characteristic.

Key Words impact strength, optical analysis, photoelasticity,

automotive interiors, natural fibers, fiber reinforcedpolymer composites

IntroductionFiber reinforced composites with thermoplastic

matrices have successfully proven their high qualitiesin various fields of technical application. Apart fromconventional fiber materials like aramide, kevlar orglass fibers natural fibers such as hemp or flax areincreasingly applied for reinforcement.

One of the major field of application for natural fiber rein-forced composites can be found in the automotive industry.Amounting to a total of only 4.000 to 5.000 tons in 1996, theuse of natural fibers in the European automotive industry has

Improving the Impact StrengthOf Natural Fiber Reinforced Composites By SpecificallyDesigned Material and ProcessParametersBy Dieter H. Mueller, Bremer Institut für Betriebstechnik und angewandte Arbeitswissenschaft an derUniversität Bremen (BIBA), Bremen, Germany and Andreas Krobjilowski, Bremer Institut fürKonstruktionstechnik (BIK), Universität Bremen, Fachbereich 04, Bremen, Germany

ORIGINAL PAPER/PEER-REVIEWED

31 INJ Winter 2004

0

25

50

75

100

0 1 2 3 4 5

Cotton

]10[mod' 3?mgravitySpecific

ulussYoung

Jute

E-Glass

FlaxHemp

Sisal

KenafRamie

Figure 1TENSILE STRENGTH AND YOUNG’S MODULUS

RELATED TO SPECIFIC GRAVITY (ρ*g) FOR DIFFERENT PLANT FIBERS COMPARED TO

SYNTHETIC FIBERS

quintuplicated with more than 28.000 tons in the year 2000. Inthe medium term, a total use of more than 100.000 tons can beexpected in this area. Currently, an average of 5 to 10 kg of nat-ural fibers is incorporated in every European passenger carwith interiors parts such as door trim panels or trunk liner asthe main fields of application [1,2].

Besides ecological considerations several technical aspectspromote the replacement of glass fibers by plant fibers.Considerably in the first place are the good mechanical prop-erties of most plant fibers coupled with a low density.Therefore, plant fibers offer a high potential for an outstand-ing reinforcement in lightweight structures (Fig. 1).

Another very important aspect regarding the replacementof glass fibers by natural fibers as the reinforcing componentin thermoplastic composites is the distinctive improvement incrash behavior. It can be assumed that that automotive interi-ors with a reinforcement of natural fibers are safer than glassfiber parts, as no sharp-edged fracture surfaces occur in caseof crash.

Further advantages of a reinforcement by natural fibersresult from their high absorptivity, which creates excellentacoustics and an air cleaning effect. With respect to industrialsafety, natural fibers do not cause allergic reactions or skin irri-tations. And finally a positive image and product marketingrelated to the utilization of a renewable material should betaken into consideration.

For the generation of structural parts by compression mold-ing carded and mechanically needle-punched nonwoven fab-rics from natural and polymeric fibers are utilized. Suchhybrid fleeces can be characterized by a maximum of homo-geneity and easy handling. For the manufacture of compositeparts pre-cut nonwoven blanks are automatically taken from amagazine and traverse a preheating device (Figure 2). Theprepreg is heated up to the required processing temperature.

The temperature has to be chosen high enough to melt andplasticize the thermoplastic binder and depends strongly onthe utilized polymer. For polypropylene (PP) temperaturesbetween 200 and 250°C are required to achieve a sufficient lowviscosity. Subsequently, in a corresponding mold the plasti-cized prepreg is finally pressed to the final shape.Simultaneously, the material is cooled down, leading to a cur-ing and hardening of the binder component. The generaldesign of the described manufacturing technology has mainlybeen taken over from the processing of glass mat reinforcedthermoplastics (GMT). Specific process adaptations have pri-marily been implemented empirically. Consequently, a lack ofknowledge, e.g. referring to the selection of process variables,still exists. However, in order to utilize the mechanical capa-bilities of natural reinforced thermoplastics in a more opti-mized way, a well-adapted process technology is required.Particularly with respect to the thermal stability, the proper-ties of glass fibers and natural fibers differ significantly. Byreason of their natural origin, plant fibers respond to thermalimpact by far more labile than synthetic fibers. For naturalfibers thermal degradation processes commence at tempera-tures as low as 120°C, resulting in a decomposition of waxes.Temperatures around 180°C lead to a decomposition of pectin,temperatures of approximately 230°C have the consequence ofa decomposition of cellulose [4]. This obviously leads to theconsumption that even transient thermal impact must resultin a non-reversible reduction of fiber strength, inevitablyeffecting the mechanical properties of the later composite.However, to ensure sufficient flow capabilities of the bindercomponent throughout the molding process, temperatures ofat least the stated magnitude are required. As mentioned pre-viously, polypropylene as the most commonly used thermo-plastic binder component for natural fiber composites,requires processing temperatures of 200 to 250°C to achieve asufficient low viscosity. Determining the parameters for annew production process therefore requires a consideration ofboth ensuring the general processability on one hand andavoiding thermal damage of the fiber component on the other.

Measurement of dynamic stress by photoelastic techniqueThe mechanical properties of fiber reinforced polymer com-

posites (FRP) have been studied especially for static loading ofthe parts. The problem of determining the distribution ofstress and/or strain in a two-dimensional composite specimensubjected to an impact type of loading is difficult to solvemathematically, thus, experimental means are essential andsometimes indispensable. Reinforced composites are mainlyused when high accelerations take place. Then the compositesoffer a high potential to reduce the kinetic energy. The rein-forcement is a major parameter to reduce the material stressnot only at static load but also when an impact loading takesplace.

Significant differences occur regarding the stress of static,quasi-static and impact loading. Impact is effected when thecontact time is short compared to the time of wave runthrough. If impact loading is subjected to a structure it is accel-erated with the consequence that interia forces are produced

32 INJ Winter 2004

Figure 2COMPRESSION MOLDING PLANT FOR THE

PROCESSING OF NATURAL FIBER REINFORCED THERMOPLASTICS

[ORIGIN: HEIDEL GMBH, GERMANY]

and therefore there is no linearity between the impact forceand the stress. The experimental simulation of an impactrequires the control of three characteristics: contact time,impulse and impact force F(t). The experimental setup for themeasurement of dynamic stress is derived from the reflectionpolariscope with single-flash-technique. The impact testdevice for the measurement of dynamic stress consists of theloading apparatus and the photoelastic element set-up. Figure3 illustrates this arrangement schematically.

To create impact loading a steel projectile is accelerated by

pressurized air along a barrel. The projectiletransmits it’s impulse to the specimenthrough an anvil (Figure 4) The use of ananvil avoids damaging the composite at thecontact point and enables the employmentof a piezoelectric force sensor. Within theloading apparatus several types of speci-men (beams, discs, plates) can be attachedwith different boundary conditions. Thus, itis possible to investigate first the ‘earlystages’ of the wave propagation (unlessreflection occurs at the point of interest ofthe specimen) and second the influence ofreflection and superposition of the propa-gating waves with respect to differentboundaries and/or profile variations. Figure4 shows a plate clamped at two oppositeedges while the remaining edges are free orsimply supported.

The testing device can develop a deeperunderstanding of the behavior of FRP com-posites after impact loadings and thus sup-port optimizations in the design and compo-sition of such materials. Within this field thefollowing applications are attainable:

1. Determination of the elastic properties ofvisco-elastic composites and photoelasticcoating materials due to high strain rates [6]

2. Location of artificial cracks within alu-minum specimen by analysis of impactstrains [7, 8]

3. Surveying and analyzing strain wavepropagation within FRP composite discs [9]

4. Investigation of the stress concentrationwithin impact-loaded orthotropic composites[10]

5. Surveying of bending waves within fiberreinforced composite plates due to transver-sal impact loadings

The latest research has involved the inves-tigation of bending wave propagation inFRP plates. Using circularly polarized lightthe isoclinic fringes disappear. The photoe-lastic images contain pure isochromaticpatterns of the wave propagation. Theisochromatic images shown in Figure 5 and

Figure 6 depict the propagation of bending waves within arectangular plate (200 * 200 * 3) mm made of a unidirection-al glass-fiber-reinforced epoxy resin. The orientation of thefibers is 60 degrees to the vertical image axis. The specimenwere loaded by central transverse im-pacts. The surveyingof the wave propagation with plane polarized light givesiso-chromatics and isoclinics de-pendent upon the angle ofthe polarization axis. When the axis of plane polarization ischanged while the loading parameters and exposure timeremain constant, a set of isoclinic fringe patterns is obtained

33 INJ Winter 2004

Figure 3IMPACT TEST DEVICE [5]

Figure 4SPECIMEN ATTACHMENT [5]

which provides the spatial distribution of the principalstrain directions.

Impact of material and process parameters on the impactbehavior of natural fiber reinforced composites

Investigations and experimental designExperimental investigations employed carded and mechan-

ically needle-punched nonwoven fabrics made of natural andpolymeric fibers. Flax, hemp and kenaf were used as naturalfibers, whereas polypropylene represented the binder fiber forall hybrid fleeces. Furthermore, the blending ratio was varied

in terms from natural/polymericfiber 50/50 to 70/30. Finallyglass fiber/polypropylene matswere used as a reference materi-al. The weight of the prepregsranged from 1600 to 2400 g/m2.

Samples were produced in lab-oratory scales, varying the leveland duration of thermal impact.The manufacture of compositesamples from prepregs tookplace in two main steps. At firstthe prepregs with thermoplasticcomponent were heated up andpre-compacted under slight pres-sure, melting and plasticizing thethermoplastic binder material.Following, these softened semi-products were consolidated totwo-dimensional plane sheets ina corresponding mold.

Influence of the processing con-ditions on the impact behavior

Figure 7 shows the influence of the process-ing temperature on the impact strength of thedifferent composite materials reinforced bynatural fibers. The processing temperaturedescribes the temperature at which the pre-heating process was interrupted. It was mea-sured by thermocouples placed in the centerlayer of the prepregs.

All composites show, regardless of theemployed type of fiber, a comparable perfor-mance with a maximum impact strength inthe medium temperatures range. For lowerand higher processing temperatures a moreor less distinctive decline can be observed.This can be ascribed to the interrelation oftwo effects opposing each other. An increas-ing processing temperature leads to lowerviscosity of the binder component and thus

to improved flow capabilities. This results in an improvedfiber embedment during consolidation and therefore higherstability of the composite. However, at the same time theincline of temperature leads to progressive decomposition ofthe reinforcing fiber, resulting in a decrease of fiber strength.

Furthermore a modification of the fiber surface caused bythermal decomposition, leading to a deterioration of adhesionbetween fiber and matrix, can also be assumed.

In the lower temperature range such decompositionprocesses are not very distinctive compared to the surplus ofmechanical properties due to progressive fiber embedment.Thus, a continuous growth in impact strength can be

34 INJ Winter 2004

Figure 5ISOCHROMATIC FRINGE PATTERN OF THE BENDING WAVE

PROPAGATION WITHIN A GRP PLATE

Figure 6ISOCHROMATIC FIGURES

observed when increasing the temperature from 180°C.Finally, a maximum is reached around a processing tempera-ture of approximately 220°C. However, for a further continu-ing increase of temperature, fiber fatigue due to thermaldecomposition becomes the predominant effect compared toa further improvement in fiber embedment, resulting in adecrease of the composite’s properties.

In the contrary, for composites with a reinforcement of glassfibers, increasing values of impact strength can be observedover the entire temperature range (Figure 8). As glass fibersendure thermal impact in the regarded temperature rangewithout any indication of decomposition, the increasingly

lower viscosity of the polymeric component leads to aprogressive improving fiber embedment and thereforehigher mechanical properties. Similar observations canbe made regarding other mechanical properties suchas tensile or bending strength or stiffness.

When increasing the share of reinforcing fiber in thecomposite, the maximum of impact strength moves tohigher temperatures (Figure 9). This phenomena can befound also for the tensile of the composites and there-fore supports the assumed superposition of the con-trary effects. On one hand an increasing share of nat-ural fibers necessarily stands for less binder material.Furthermore, due to their comparatively large diame-ter and rough surface morphologies natural fiberscounteract to a free flow of the binder component dur-ing consolidation. Therefore a lower binder viscosity isrequired to achieve a reasonable fiber embedment forhigher shares of natural fiber. Thus, the peak value ofimpact strength is reached at higher processing tem-peratures and the strengthening influence of improvedfiber embedment predominates longer over the weak-ening thermal decomposition.

Not only the level of thermal impact, represented bythe absolute processing temperature, was found to beresponsible for a decline in fiber strength but also theduration as shown in Figure 10. Most of the productsused for example in the automotive industry, have ablending ratio of 50/50 for natural fibers and 30/70 forglass fibers. Investigations [11], [12] demonstrate thatthese rations generate the optimal tensile strengths.For the composite reinforced by natural fibers the mea-sured impact strength depends strongly on the dura-tion of the thermal impact, whereas the most signifi-cant loss in impact strength can be observed for dura-tions comparable to industrial standard cycle times. Incontrary, no dependency can be observed for glassfiber reinforcement over the same range of time.

Concluding it has to be recorded that both, the levelof thermal impact, represented by the absolute pro-cessing temperature, as well as the duration, repre-sented by the preheating time, have a strong influenceon the decomposition of natural fibers and thereforeon the mechanical properties of natural fiber rein-

forced composites.

Influence of the fiber characteristics on the impact behaviorFigure 11 summarizes the maximum mechanical properties

of all investigated composite materials employing an opti-mized processing as described above. The given values arepercent specifications, related to the particular maximumwhich is also given on each axis.

Obviously flax fibers offer the highest reinforcing potentialof all investigated natural fibers. As it is evident in Figure 11this superiority is most distinctive regarding the tensile andimpact strength of the composites.

35 INJ Winter 2004

Figure 7IMPACT STRENGTH VS. PROCESSING TEMPERA-

TURE FOR COMPOSITE REINFORCED BY DIFFERENTNATURAL FIBERS, ALL COMPOSITES NATURAL

FIBER/PP 50/50, 1600 G/M2

Figure 8IMPACT STRENGTH VS. PROCESSING TEMPERA-TURE FOR COMPOSITE GLASS FIBER/PP 30/70,

1600 G/M2

To verify this assumption, Figure 12 compares the tensilestrength data for composite materials basing on PP and dif-ferent fiber materials. Regardless of the employed type, theincorporation of a reinforcing component results in anincrease of the composite’s tensile strength compared to thepure PP reference sample. Furthermore it is obvious that com-posite materials with a reinforcement of natural fibers offerthe same dimension of tensile strength as E-glass reinforcedcomposites. The composites basing on flax and hemp evenshow superior tensile strength values. Even though the fibertensile strength values were basically comparable at least forflax and hemp fibers (Flax: 45 cN/tex, Hemp 43 cN/tex,

Kenaf 35 cN/tex), composites basing on flaxfibers show higher tensile strength values thanthose basing on hemp or kenaf fibers. From thatit can be concluded that flax offers higher rein-forcing properties than hemp and kenaf. A rea-son for this can be seen in the fiber morphology.Flax fibers show a higher fineness and alsomore unique fiber diameter distribution com-pared to hemp or kenaf. Especially the latterusually can be characterized by wide rangingdiameter distributions and comparatively largediameters. A high fiber fineness should result inbetter fiber embedment during compressionmolding and therefore higher mechanical com-posite properties. Furthermore a higher fiberfineness should lead to an improvement in theratio between surface and volume and there-fore an increased contact surface between fiberand polymeric matrix. Furthermore the utiliza-tion of a fiber with less diameter increases thenumber of actually embedded fibers whichreduces the concentration of peak stresses atthe end of the fibers. To verify these presump-tions Fig. 12 also shows the average fiber diam-eter of the tested natural fibers. It is obviousthat a utilization of fibers with a higher finenessindeed leads to improved composite proper-ties. Fig. 12 also presents the influence of theshare of natural fiber material in the composite.Obviously, a maximum in tensile strength isreached with a share of 60% of flax fiberswhereas higher or lower shares result in adecrease in tensile strength. Similar results canbe attained for the stiffness of the compositematerials, i.e. the modulus of elasticity in ten-sion.

Particularly with respect to impact strengththe fiber fineness plays a key role for the com-posite properties. Tests with glass fiber rein-forced thermoplastic composite could provethat an increase in fiber diameter leads toreduced impact strength [11]. A fiber with lessdiameter offers the advantage of an increasingabsorption for impact energy. Furthermore the

capability of fiber-pull-out is also increased. Figure 13 displays the maximum values in impact strength

for the different composite materials. Also given are the aver-age diameter and the elongation at break of the fiber material.As expected composites with fine reinforcing natural fibersoffer the highest impact strength. At the same time also highfiber elongation capabilities lead to high impact strength.

The impact strength of composite materials reinforced bynatural fibers is often considered to be inferior to a reinforce-ment with glass fibers. Figure 13 displays that basing on anoptimized manufacture referring to thermal processing condi-tions as described above even the impact strength can reach

36 INJ Winter 2004

Figure 9RIMPACT STRENGTH VS. PROCESSING TEMPERATUREFOR COMPOSITE WITH DIFFERENT SHARES OF REIN-

FORCING FIBER, ALL COMPOSITES FLAX/PP, 1600 G/M2

Figure 10IMPACT STRENGTH VS. PREHEATING TIME FOR DIFFER-

ENT COMPOSITES, ALL COMPOSITES NATURAL FIBER/PP50/50, 1600 G/M2, EXCEPTION: GLASS FIBER/PP 30/70,

1600 G/M2

values well comparable to those of glass fiber composites. Forhigh impact properties it has to be noted, that a slightly weak-er adhesion between fiber and polymer should result in ahigher degradation of impact energy, supporting fiber-pull-out. Good adhesion on the contrary results in abrupt fiberfracture with a minor energy degradation.Summary

The investigations could prove a strong influence of thethermal process conditions during compression molding onthe impact strength of natural fiber reinforced composites.

Regardless of the employed natural fiber a peak of in impactstrength was found at processing temperatures around 220°Cto 240°C. Higher temperatures as well as lower values result-ed in lower mechanical values. This observation may beexplained by a thermal decomposition of the fiber compo-nent. Similar experiments with glass fiber reinforced compos-ites showed no thermal impact on the material properties.

Furthermore a strong impact of different material parame-ters on the composite’s properties exists. Especially the fiberfineness has a strong impact on the composite’s mechanicaldata. A higher fineness results in an improved fiber length todiameter relation and an increased contact surface betweenfiber and matrix. Furthermore it leads to an increased amountof fibers in the composite and a decrease in stress concentra-tion

References1. Bledzki, A. K., Gassan, J., Composites reinforced with cel-

lulose based fibres, Progress in Polymer Science, 24, (1999),pages 221-274

2. Karus, M., Kaup, M., Lohmeyer, D., Studie zur Markt-und Preissituation bei Naturfasern, Nova Institut,Hürth/Germany, March 2000

3. Kaup, M., Karus, M., AktuelleMarktübersicht Naturfasern für technischeTextilien in der EU, Proceedings, 3.International Symposium “Werkstoffe ausnachwachsenden Rohstoffen,”Erfurt/Germany, 5./6. September 20014. Knothe, J., Fölster, Th.,Naturfaserverstärkte Fahrzeugteile,Kunststoffe 87 (1997), Carl Hanser Verlag,Germany, pages 1148-11525. Franz, Th., Experimentelle Untersuchungvon impactbelasteten versteiften bzw.gekerbten Platten ausFaserverbundwerkstoffen, Dissertation,University of Bremen (Germany), 19986. Tenzler, A., Müller, D.H., Ermittlungdynamischer Werkstoffkennwerte an FVWmit der Spannungsoptik, VDI Werkstofftag,München, VDI Bericht 852, 1991, pages 759-7667. Elmer, K.-H., Meltzer, G., Tenzler, A.,Weber, Th., Detektion von Rissen mit mech-anischen Stoßwellen, VDI-Z 136(7/8), 1994,pages 85-87

8. Franz, Th. & D.H. Müller 1995. Localization of crackswithin construction parts by analysis of strain waves, Proc.Annual BSSM Conf., Sheffield, UK, September 5-7th, 1995,pages 22-24

9. Müller, D.H., Jüptner, W., Franz, Th., Geldmacher J.,Tenzler, A., Untersuchung der Stoßwellenausbreitung inFaserverbundwerkstoffen mittels dynamischer Spannungs-optik und holografischer Interferometrie, Forschung imIngenieurwesen - Engineering Research 62(7/8), 1996, pages195-213

37 INJ Winter 2004

Figure 11INFLUENCE OF THE REINFORCING FIBER ON

THE COMPOSITE PROPERTIES, PERCENTSPECIFICATION RELATED TO SPECIFIC MAXI-

MUM, OPTIMIZED PROCESSING, COMPOS-ITES 1600 G/M2

Figure 12TENSILE STRENGTH FOR COMPOSITES BASING ON DIF-

FERENT FIBER MATERIALS AND PURE PP IN COMPARISONTO TENSILE STRENGTH FOR SEPARATE FIBERS (SHADED

BARS) AND AVERAGE FIBER DIAMETER (OFDA), OPTI-MIZED PROCESSING, ALL COMPOSITES 1600 G/M2

10. Tenzler, A., Entwicklung einer Methode zur Ermittlungvon Kerbwirkungen an impactbeanspruchtenFaserverbundwerkstoffen, Fortschr.-Ber. VDI Reihe 18 Nr. 158,Düsseldorf: VDI-Verlag, 1994.

11. N.N., Verstärkte Thermoplaste: Einfluss desGlasfilamentdurchmessers, Vetrotex Glasfilament Report,1995/1, pages 8-9

12. Krobjilowski, A., Prozess- und Produktoptimierungbeim Formpressen naturfaserverstärkter thermoplastischerVerbundwerkstoffe, PhD-Thesis University Bremen,27.06.2003 — INJ

38 INJ Winter 2004

Figure 13IMPACT STRENGTH FOR COMPOSITES BASING ON DIFFER-ENT FIBER MATERIALS AND PURE PP IN COMPARISON TOELONGATION AT BREAK FOR SEPARATE FIBERS (SHADED

BARS) AND AVERAGE FIBER DIAMETER (OFDA), OPTIMIZEDPROCESSING, ALL COMPOSITES 1600 G/M2