Natural Fiber Composites Based on Flax - Matrix Effects

Transcription

1 International Scientific Colloquium Modelling for Saving Resources Riga, May 17-18, 2001 Natural Fiber Composites Based on Flax - Matrix Effects R. Joffe, L. Wallström and L.A. Berglund Abstract The main objective of this study is to select the best resin/fiber combination considering matrix/fiber compatibility, stiffness, strength and fracture toughness of the natural fiber composites (NFC). These properties are measured and compared with respect to the reference material glass mat thermoplastic (GMT). Five different thermoset resins were used to manufacture NFC by resin transfer molding. GMT with polypropylene as a matrix was manufactured by compression molding. This investigation shows that NFC have mechanical properties as high as GMT or even higher in some cases. Such good mechanical properties in combination with light weight makes use of NFC very attractive for automotive industry. The composite performance is analyzed in terms of constituent properties and interface quality. Introduction Automotive industries in Europe show large interest in NFC that can be used in loadbearing elements of cars. These NFC compete with and substitute GMT in wide-spread use today. The advantages of NFC are related to environmental aspects and low weight. Also thermoset matrices are of interest where compression molding or resin transfer molding (RTM) are suitable processes. In order to develop high performance NFC, automotive industry and research institutions in Europe are collaborating in the EU GROWTH Project ECOFINA. The study presented here is a result from this project. The main objective of this study is to choose the best resins, considering matrix/fiber compatibility, stiffness, strength and fracture toughness of the NFC. These properties will be compared with respect to the reference material - GMT. Experimental investigations of different resins reinforced with a commercially available flax fiber mat are performed. Five different thermoset resins (three polyesters, epoxy and vinylester) are used to manufacture NFC by the RTM process. Along with the strong industrial orientation of this study, there are some aspects that need deeper scientific investigation. Although resins have very different properties, many composite properties show only small variation with different resins. Apparently the fibers and the fiber/matrix interface [1,2] control several properties. At this stage of the presented study the problem is not solved, only described. However, future work in this project is planned on modeling/prediction of fracture toughness of NFC by using the bridging law concept [3,4]. This involves measurement of a characteristic bridging law, where the strength of fiber-matrix interface influences its shape [5,6]. 54

2 1. Experimental 1.1 Composites Manufacturing and Testing Composites were manufactured by using the following thermosets: three polyesters, Palatal E (A) by DSM; Norpol (B) by Reichhold; Norpol (D) by Reichhold, vinylester XZ92485 (E) by DOW, epoxy XZ92474/XZ92415 (C) by DOW. Also thermoplastic resin was used in GMT - standard polypropylene (F). Designation in brackets A-F is introduced for convenience of data presentation. Plates (thickness 3 mm) of NFC were manufactured by RTM and postcured for 5h at 80 C after gelation. Commercially available nonwoven mat of mechanically separated flax fiber (FF) produced by Muhlmaier was used as a reinforcement, resins A-E were used as matrix material. GMT plate (thickness 3.4 mm) was produced by compression molding, commercially available glass fiber (GF) mat with fiber length of 25 mm was used as reinforcement and polypropylene (PP) was used as a matrix material. Volume fraction of fibers was 30% (36% by weight) for NFC and 20% (41% by weight) for GMT. Such proportions gave almost equal densities of composites, approximately 1.25 g/cm 3 for NFC (density of FF 1.50 g/cm 3 ) and approximately 1.30 g/cm 3 for GMT (density of GF 2.59 g/cm 3 ). Flat rectangular specimens were cut from composite plates with dimensions according to ASTM standard D a. CT specimens had the same dimensions as in case of neat resins, except thickness. Tensile fracture toughness tests were done according to ASTM standards D a [7] and D a [8] respectively. Both tests were stroke controlled. Tensile test was performed at loading rate of 2 mm/min and fracture toughness test at 10 mm/min, 5-6 specimens for each test were used. Tensile modulus of composites was measured in linear part of the unloading stress-strain curve in strain interval of %. All other measurements are done according to procedure described in standards Neat Resins Manufacturing and Testing Three plates of neat resin C, D and E of thickness 2.4 mm also were made. Plates were postcured for 5h at 80 C after gelation. Dogbone shaped specimens (Type II) for tensile tests were manufactured according to ASTM standard D [9]. Compact tension (CT) specimens (width W = 20 mm), to measure fracture toughness, were produced according to ASTM standard D a. Tensile and fracture toughness tests were done according to ASTM standards D and D a respectively. Both tests were stroke controlled. Tensile test was performed at loading rate of 5 mm/min and fracture toughness test at 10 mm/min, 5-6 specimens for each test were used. All measurements are done according to procedure described in standards. 2. Results and Discussion 2.1. Tensile Properties of the Composite Average values and standard deviation of tensile modulus, strength and strain of failure for composites are presented in Tab. 1. In order to plot all these data in one graph all properties of NFC are normalized with respect to the corresponding properties of GMT and presented in Fig. 1. These data show that all tested NFC have tensile modulus higher than GMT and the difference for materials B-E is rather significant (50-100%). Strength of NFC is also considerably higher (10-40%) than for GMT material, except for batch A. Tab. 1. Tensile properties of composites. Average values with standard deviation. 55

3 Material E (GPa) σ max (MPa) ε max (%) A 6.11 ± ± ± 0.14 B 8.35 ± ± ± 0.11 C ± ± ± 0.06 D ± ± ± 0.11 E 9.76 ± ± ± 0.08 F 5.32 ± ± ± 0.23 Tensile Properties ofcom posites N orm alized tensile properties A B C D E F Batch Y oung's m odulus Strength Failure strain Fig. 1. Tensile properties of composites normalised with respect to GMT properties (F). However, values of failure strain of all NFC are lower (10-35%) than for GMT. Therefore, in order to choose NFC with best performance only tensile modulus and strength can be considered as an optimisation parameters. The most interesting materials are NFC with the following matrices: epoxy C, polyester D and vinylester E Fracture Properties of NFC and Neat Resins Similar selection procedure can be done based on measured fracture toughness K IC of NFC and neat resin. Experimental data are presented in Tab. 2 and normalized values of average K IC for NFC and neat resin with standard deviation are presented in Fig. 2. These data are also normalized with respect to fracture toughness of PP and GMT for neat resin and NFC respectively. Pure PP was not tested but mechanical properties of PP are available in literature [10]. However, fracture toughness of PP found in different sources varies depending on PPgrade tested, therefore, an average value (from those found in literature) of 3 MPa m 1/2 is assumed. These data show that GMT has highest value of fracture toughness (see Fig. 2 series NFC norm. w. GMT ). NFC materials B-C, considering data scatter, showed almost equal values of K IC. This means that the deciding parameters in material selection are still Young s modulus and strength of NFC, which means that materials of interest are C, D and E. Tab. 2. Fracture toughness of NFC and neat resin. Average values with standard deviation. 56

4 Material NFC: K IC MPa m 1/2 Neat resin: K IC MPa m 1/2 A 3.17 ± 0.41 not tested B 4.53 ± 0.30 not tested C 4.45 ± ± 0.56 D 5.01 ± ± 0.15 E 5.16 ± ± 0.65 F 6.86 ± from literature Fracture T oughness ofnfc and NeatResin N orm alized fract.toughness NFC norm.w.gm T N eat resin norm.w.pp NFC norm.w.neat resin A B C D E F Batch Fig. 2. Fracture toughness of neat resin and NFC Failure Mechanisms in Flax Fiber Composite Mechanical properties of three selected NFC are very similar to each other (see Fig. 1 and Fig. 2). However, if properties of resins that are used as matrix in these composites are compared, grading of material by mechanical properties is rather different. Fracture toughness of resin D is two times lower than for C and E (Fig. 2 series Neat resin norm. w. PP ). If increase of fracture toughness due to reinforcement is compared (see Fig. 2 series NFC norm. w. neat resin ) it is clear that in case of material D (4.2 times increase) fracture toughness increased much more than for C and E (1.7 times). Different level of effect of reinforcement is even more pronounced if tensile properties of neat resin is analysed. Tensile properties of neat resin obtained from manufacturers data sheets together with measured values (for material C, D and E) are presented in Tab. 3. The same data but normalised with respect to corresponding tensile properties of PP are presented in Fig. 3. These data show that difference of properties of neat resin is very large but there is almost no difference when these resins are used in composites. Material A is very good example, strain at failure (Fig. 3) of neat resin is almost 10 times higher than for all other resins (except E where difference is 5 times). However, NFC based on this resin has lowest fracture toughness, and strain at failure just slightly higher than for other NFC. Similarly, resin D indicates lowest values of fracture toughness and strain at failure in case of neat resin but equivalent or even higher compare to other NFC properties in case of composite. Failure of the composite is obviously not controlled by the matrix. On the other hand fibers are the same in all tested composites and one can expect that material grading by properties of NFC 57

5 would follow grading of neat resin, which is not the case. Most likely this is because fibers do not carry the same load in different NFC due to different load transfer mechanisms during the failure process. Tab. 3.Tensile properties of neat resin. Average values with standard deviation. Mat. Manufacturer data sheet Experimental E (GPa) σ max (MPa) ε max (%) E (GPa) σ max (MPa) ε max (%) A B C ± ± ± 1.3 D ± ± ± 0.1 E ± ± ± 0.8 F * * - from literature [10] Tensile Properties ofresins N orm alized tensile properties >9 Y oung's m odulus (m anufact.) Strength (m anufact.) Failure strain (m anufact.) Y oung's m odulus (m easured) Strength (m easured) Failure strain (m easured) A B C D E F Batch Fig. 3. Tensile properties of neat resin normalised with respect to properties of PP (F). This indicates that the fiber-matrix interface differs for different resin/fiber combinations. Therefore, mechanisms that determine fracture properties of NFC are fiber debonding, fiber pull-out and resulting crack bridging. For example, if in material D the interfacial strength is very low this would lead to extensive fiber debonding and pull-out when crack is formed. Pull-out dissipates large amounts of energy. Undoubtedly this would delay crack propagation, which leads to a more ductile behavior of the composite compared with the brittle behavior of the neat resin. Such behavior can be modeled using crack bridging concepts, which requires measurement of bridging law and would take into account interfacial strength of fiber-matrix interface. 58

6 Conclusions This experimental study showed that NFC based on thermoset resins and flax fibers have mechanical properties that are comparable (or even higher) with GMT. This makes such material very attractive for automotive industry where GMT is traditionally used. Three materials out of five NFC were selected based on Young s modulus and strength values. These materials are based on epoxy, vinylester and polyester resin. Such wide range of resins makes it possible to select best material based not only on optimization of mechanical properties but also on other important issues, such as impact on environment or price for instance. Experimental results indicated that failure mechanism in NFC is rather complicated and must be further investigated. This investigation should include study of fiber/matrix interface and results can be used in prediction of fracture toughness of NFC by application of crack bridging concepts. More knowledge on interface properties will allow designing composites with better performance. Acknowledgments Authors would like to thank Mr. Patrik Fernberg for fruitful discussions concerning crack bridging methods. References [1] Chamis C.C.: Mechanics of load transfer at the interface. Composite Materials, Vol. 6, 1974, pp [2] Pitkethly M.J.: The use of interfacial test methods in composite materials development. Fiber, Matrix, and Interface Properties, ASTM STP 1290, C.J. Spragg and L.T. Drzal, Eds., ASTM, 1996, pp [3] Bao G., Suo Z.: Remarks on crack-bridging concepts. Applied Mechanics Review, Vol. 45, August 1992, pp [4] Lindhagen J., Berglund L.A.: Application of bridging-law concepts to short-fibre composties Part 1: DCB test procedures for bridging law and fracture energy. Composites Science and Technology, Vol. 60, 2000, pp [5] Hutchinson J.W., Jensen H.M.: Models of fiber debonding and pullout in brittle composites with friction. Mechanics of Materials, Vol. 9, 1990, pp [6] McCartney L.N.: Mechanics of matrix cracking in brittle-matrix fibre-reinforced composites. Proc. R. Soc. Lond., Vol. A409, 1987, pp [7] ASTM D a: Standard test method for tensile properties of polymer matrix composite material. [8] ASTM D a: Standard test methods for plain-strain fracture toughness and energy release rate of plastic materials. [9] ASMT D638-95: Standard test method for tensile properties of plastics. [10] Holmberg J.A. and Nilsson P.: Initial testing of Symalit GM40PP and pure PP. SICOMP Technical Report , 1998, Swedish Institute of Composites, Piteå, Sweden. Authors Dr. Roberts Joffe Research Associate Div of Polymer Engineering Luleå University of Technology S Luleå, Sweden Roberts.Joffe@mb.luth.se Dr. Lennart Wallström Senior Lecturer Div of Polymer Engineering Luleå University of Technology S Luleå, Sweden Lennart.Wallstrom@mb.luth.se Dr. Lars Berglund Professor Div of Polymer Engineering Luleå University of Technology S Luleå, Sweden Lars.Berglund@mb.luth.se 59