2 - Institut für Verbundwerkstoffe GmbH, Universität Kaiserslautern, D Kaiserslautern, Germany

Transcription

1 A Comparison of the Mechanical Behaviour of Weft-Knitted Glass and Carbon Fiber Fabric-Reinforced Polyamide-12 A Comparison of the Mechanical Behaviour of Weft-Knitted Glass and Carbon Fiber Fabric-Reinforced Polyamide-12 T. Czigány *1 and J. Karger-Kocsis Department of Polymer Engineering and Textile Technology, Faculty of Mechanical Engineering, Budapest University of Technology and Economics, H-1111 Budapest, Müegyetem rkp. 3, Hungary ( czigany@eik.bme.hu) 2 - Institut für Verbundwerkstoffe GmbH, Universität Kaiserslautern, D Kaiserslautern, Germany Received: 18th May 2001; Accepted: 31st August 2001 SUMMARY The mechanical response of knitted glass- (GF) and carbon fiber (CF) fabric-reinforced polyamide-12 (PA-12) composites was compared both under static (tensile, flexural) and dynamic (perforation impact) conditions. The GF- and CF-content of the composites produced by hot pressing of 8 layers of planar weft knits containing staple reinforcing GF or CF and PA-12 fibers was practically the same, viz ca. 50 vol%. Dynamicmechanical thermoanalysis (DMTA) showed a strong stiffness anisotropy: the stiffness of the composites in the wale-direction (W) was markedly higher than in the coursedirection (C). The same strong anisotropy was also found for the static tensile and flexural characteristics (stiffness and strength). The mechanical anistropy and the large difference between the stiffness and strength due to tensile and flexural loading were traced to the structure of the knits ( unbalanced reinforcement) and their non-planar structure in the composite sheets. The mechanical performance of the knitted CFreinforced composite was superior to the GF-reinforced version under dynamic perforation impact conditions as well. 1. INTRODUCTION Nowadays considerable research and development activity is devoted to knitted fabric-reinforced thermoplastic composites [1-2]. Note that the nonconsolidated plain (planar) knits can easily be produced from commingled yarns or by co-knitting (i.e. an in-situ combination of reinforcing and hot pressing of the matrix giving fibers) [1-4]. Such nonconsolidated knits exhibit excellent drapability and net shape processability. They possess outstanding deep-drawability even if consolidated knitted fabricreinforced thermoplastic sheets are used as the raw materials. The wetting out of the reinforcing fibers by the polymer melt is very good as the matrixforming fibers are well dispersed within the yarn or tow used for knitting. Drawbacks with knitted reinforcements are due their interlooped, coherent structure. Because of the interlooping, the maximum * author to whom correspondence should be addressed fiber packing is lower than in woven fabric-reinforced composites. Further, the severe bending of the yarn may cause fiber fracture and result in poor mechanical properties. The coherent structure of the knit means that its stretching in one direction is inevitably accompanied by a contraction in the other. Large loop size and heavy tows are the major sources of problems with knitted reinforcements. The distribution of the reinforcement in a matrix polymer containing one single layer of plain knit is highly inhomogeneous as it follows an all (only reinforcing yarn) - nil (only matrix) pattern. The reinforcement homogenization can be achieved by stacking several knit layers, the minimum number of which depends on the loop size and yarn fineness [5]. Irrespective of the obvious benefits of using knitted fabrics for thermoplastic reinforcements, the majority of the relevant publications are related to thermoset-based composites, according to a recent review [6]. The objective of this study was to compare the mechanical performance of plain weft knitted GF- and CF-fabric Polymers & Polymer Composites, Vol. 9, No. 8,

2 T. Czigány and J. Karger-Kocsis reinforced polyamide-12 (PA-12) composites, at the same reinforcement content (V f 50 vol%). The composite sheets were produced by hot pressing knit layers consisting of staple glass (GF) and polyamide- 12 (PA-12), and carbon fiber (CF) and PA-12 fibers, respectively. Note that staple fiber yarns composed of GF, CF and PA-12 were recently introduced by EMS Chemie (Domat, Switzerland) [7]. One of the major advantages of commingled staple fibers is that the reinforcement content can be set in a broader range than by using continuous fibers. 2.1 Materials 2. EXPERIMENTAL Weft-knitted fabrics of staple CF/PA-12 and GF/PA- 12 fiber yarns were produced on a circular knitting machine using the contrary knitting technique [8]. The staple yarn was twined with a continuous polyamide-12 yarn. The nominal GF- and CF-content was set at 50 vol%. The composite plates were manufactured by hot pressing after stacking 8 knit layers (4 double layers). The processing conditions were as follows: T=225 C and consolidating pressure p=3 MPa. These press conditions were found to be the optimum for the knits used [8]. The thickness of the pressed sheets was 1.4 and 1.8 mm for the GF/ PA-12 and CF/PA-12 materials, respectively. The knitted fabric had a loop with ca. 4 mm width (course) and ca. 8 mm height (wale). Static Tensile and 3-Point Bending Tests The static tests were carried out at room temperature and 1 mm/min crosshead speed on a Zwick 1474 universal testing machine. Tensile tests followed the EN 61 standard; the size of the rectangular specimens with end tabs was 160 mm x 25 mm x thickness (clamped length: 90 mm). The 3-point bending test was performed according to the EN 63 standard, the size of the specimens was 70 mm x 25 mm x thickness (span length: 16x thickness). According to our previous study using acoustic emission to determine the damage zone size [9] it is necessary to make the specimens width larger than four times the representative elementary unit cell (cf. Figure 1). Figure 1 Structure of single layer of the plain weft knitted fabric along with the size of the elementary unit cell. The knitted fabric had a loop with ca. 4 mm width and ca. 8 mm height. 2.2 Specimen Preparation and Testing Static and dynamic mechanical tests were carried out to assess the mechanical performance of the weft knitted GF- and CF-reinforced PA-12 composites. Specimens of rectangular shape were cut from the sheets in both wale (W) and course-directions (C) by sawing. Five specimens were tested in each direction at room temperature (RT). Dynamic-Mechanical Thermoanalysis (DMTA) Specimens of 70 x 13 x thickness mm 3 were subjected to load-controlled sinusoidal tensile loading at a frequency of 10 Hz in a DMTA device (Eplexor 150N, Gabo Qualimeter, Ahlden, Germany). The complex tensile modulus (E* = E + ie ) and its components (the loss modulus, E and storage modulus, E ) together with the mechanical loss factor (tan δ = E / E ) were monitored from -100 C to +200 C at 1 C/min heating rate. The static and oscillating load components were 80 and ±50 N, respectively. Instrumented Falling Weight Impact (IFWI) Quadratic plates of ca 70x70 mm 2 surface were subjected to IFWI. IFWI of the plaques was carried out using a Dartvis device connected to an AFS-MK4 data acquisition module from Ceast (Pianezza, Italy). The inner diameter of the circular clamping unit was 40 mm, whereas the diameter of the impactor was 20 mm. This testing set-up reflects the interfacial changes more accurately because of the presence of shear stresses accommodated in the sheet during loading. 492 Polymers & Polymer Composites, Vol. 9, No. 8, 2001

3 A Comparison of the Mechanical Behaviour of Weft-Knitted Glass and Carbon Fiber Fabric-Reinforced Polyamide-12 In addition, the span-to-thickness ratio was closely matched to that of the 3-point bending set-up (viz. 16 x thickness) for the sake of comparison of the related results. The load transducer built in the impactor tip allowed us to monitor the impact response (loadtime) of the materials. The speed of the dart (v) was measured just before impacting the plate by a light cell. Perforation impact was performed occurred at ambient temperature with v=10 m/s; the incident impact energy was J. 3.1 DMTA 3. RESULTS AND DISCUSSION Figures 2 and 3 show the absolute value of the complex dynamic modulus (E*) and the mechanical loss factor (tan δ) as a function of the temperature (T) both in the W- and the C-directions. The glass transition (T g ) of PA-12 can well be resolved at T 50 C especially in the tan δ versus T trace taken in the C-direction. The stiffness anisotropy (compare the course of the stiffness in the W- and C-directions) is clearly seen in Figures 2 and 3. This suggests that the unbalanced feature of the knit (i.e. the amount of knit segments oriented in the W-direction) is markedly higher than in the C-direction. This finding is in harmony with several reports on weft knitted fabric-reinforced thermoplastic composites listed in Ref. [1]. 3.2 Tensile and Flexural Response The tensile characteristics reflecting the aforementioned strong mechanical anistropy are summarized in Table 1. One can recognize that both ultimate tensile strength (σ) and E-modulus (E) are markedly higher for CF- than for GF-knit reinforced PA-12 composites. Further, the E-moduli agree fairly with those derived from the DMTA spectra (cf. Figures 2 and 3). The strong mechanical anisotropy is clearly observable also in flexural loading. Figure 4 shows characteristic force-deflection curves of the composites with GF and CF knit reinforcement in W- and C-directions, respectively. Unfortunately, these force-deflection curves cannot be directly compared as the thickness of the specimens was different. The flexural characteristics are tabulated also in Table 1. Comparing the tensile and flexural data in Table 1, one can recognize that the flexural stiffness and strength are superior to the related tensile characteristics. It was argued that this is due to the smaller volume under stress in 3-point bending, which depresses eventual effects of inhomogeneities and defects. Further, the knit reinforcement in the composites is not planar but strongly undulated. This obviously alters the through-thickness profile of the compressive and tensile stresses which is no longer smooth [1, 10]. Figure 2 Characteristic DMTA spectra (E* vs. T and tanδ vs. T traces) of knitted GF fabric-reinforced PA-12 composite in wale and course direction, respectively Polymers & Polymer Composites, Vol. 9, No. 8,

4 T. Czigány and J. Karger-Kocsis Figure 3 Characteristic DMTA spectra (E* vs. T and tanδ vs. T traces) of knitted CF fabric-reinforced PA-12 composite in wale and course direction, respectively Figure 4 Representative force-deflection curves registered in 3-point bending for knitted GF/PA-12 and CF/PA-12 composite specimens in wale (W) and course (C) directions, respectively Table 1 Tensile and flexural strength ( σ) and stiffness (E) of knitted GF- and CF-fabric-reinforced PA-12 composites in wale and course directions, respectively Composite GF/PA-12 CF/PA-12 Direction Wale Course Wale Course Tensile Parameters σ [ MPa] [GPa] 214± ± ± ± 10 Flexural Parameters E σ [ MPa] E [GPa] 17.0± ± ± ± ± ± ± ±9 25.6± ± ± ± Polymers & Polymer Composites, Vol. 9, No. 8, 2001

5 A Comparison of the Mechanical Behaviour of Weft-Knitted Glass and Carbon Fiber Fabric-Reinforced Polyamide IFWI The primary data were stored on a computer so that various plots (e.g. load or energy as a function of time or deflection) could be generated. From the IFWI fractograms the maximum bending strength (σ d ), the thickness-related maximum load (F max /t) and perforation energy (E perf /t), deflections at the load maximum and perforation (x max and x perf, respectively) the ductility index (DI) and the disc modulus (E d =E/ (1-ν 2 )) were read or computed by the related data reductions [11]: σ d 25 F = 2 t. max (1) E Eperf = max (2) t E DI = E E max Fmax max (3) 2 F d Ed = x 3 (4) t Where F/x = slope of the F-x trace, d = diameter of the disc support (40 mm of the present test), t = thickness of the sheet, E perf = perforation energy, E Fmax = energy absorbed until F max, and E max = total absorbed energy. The penetration site was studied by light microscopy. Characteristic load and energy versus deflection traces for GF/PA-12 and CF/PA-12 are displayed in Figure 5. As expected, the maximum load, and perforation impact energy along with the related deflections are markedly higher for CF than for GF knit reinforcement - cf. Table 2. This is not only due to the higher stiffness and strength of the CF but also to a strong interphase between CF and PA-12. It is supposed that the interfacial bonding was favoured by transcrystallization in CF/PA-12 [12]. Figure 5 Characteristic force-deflection and energy-deflection traces for the knitted GF/PA-12 and CF/PA-12 (Note: for sake of comparison, both the force and energy were normalized on the thickness (t) of the sheets impacted) Table 2 Perforation impact characteristics of knitted GF/PA-12 and CF/PA-12 composite sheets Composite E d ] [GPa s d ] [MPa F max / t [N/mm] X F max ] [mm X p erf ] [mm E perf / t [J/mm] DI [%] GF/PA ± ± ± ± ± ± ± 4 CF/PA ± ± ± ± ± ± ± 8 Polymers & Polymer Composites, Vol. 9, No. 8,

6 T. Czigány and J. Karger-Kocsis Figure 6 Perforation failure of a knitted GF/PA-12 sheet the flexural stiffness and strength are higher than the corresponding tensile values was explained by the difference in the stressed volume (lower under flexure and thus more sensitive to variation in reinforcement content and less sensitive to defects compared with tensile) and by the quasi three-dimensional reinforcing structure of the knit layers in the consolidated sheets. Figure 7 Perforation failure of a knitted CF/PA-12 sheet ii. The CF-reinforced composite outperformed the GF-version both under static (tensile, flexural) and dynamic (perforation impact) loading. This was traced to the high strength of the CF and good interfacial bonding between CF and PA- 12 (transcrystallization may be responsible for the latter). ACKNOWLEDGEMENTS Figures 6 and 7 illustrate the perforation failure in GF/PA-12 and CF/PA-12 composites. Note that no difference can be revealed between the GF and CF knit-reinforced composites, at least on a macroscopic level. The perforated hole is not circular but somewhat elliptical in shape with a length axis to W-direction. This is linked again to the inherent anisotropy of the knit used [13]. The outstanding resistance to perforation of the CF knitted fabric reinforced PA-12 sheet compared to the GF-version can be attributed to the higher strength of the CF and to the good interfacial bonding between CF and PA- 12. The effect of the latter has been demonstrated on the example of knitted GF-reinforced polypropylene composites having various interfacial shear strength values [14]. 4. CONCLUSIONS Based on this study performed on the static and dynamic mechanical characterization of plain weft knitted glass (GF) and carbon fiber (CF) fabric reinforced polyamide-12 (PA-12) composites produced by hot pressing from 8 layers of commingled staple yarn (reinforcement content 50 vol.%) knits, the following conclusions can be drawn: i. The static tensile and flexural parameters of the composite sheets are strongly anisotropic, because of the inherent anisotropy of the knit used ( unbalanced reinforcement). The finding that This work was supported by the Hungarian Scientific Research Fund (OTKA F ) and German Science Foundation (DFG FOR 360). T. Czigány wishes to acknowledge the Széchenyi fellowship of the Hungarian Ministry of Education. The authors would like to thank the Institute for Design and Construction Methods (ETH Zurich, Switzerland) for the production and delivery of the composite plates. REFERENCES 1 Czigány T., Karger-Kocsis J., Textile fabricreinforced thermoplastic polyester composites in Handbook of Thermoplastic Polyesters, Ed.: Fakirov S., Wiley-VCH, Weinheim, 2001, in press 2 Mayer J., Wintermantel E., Thermoforming processes for knitted-fabric-reinforced thermoplastics: New manufacturing techniques for load-bearing, anisotropic implants in Composite Sheet Forming, Ed.: Bhattacharyya D., Elsevier, Amsterdam, 1997, pp Friedrich K., Commingled yarns and their use for composites in Polypropylene: An A-Z Reference, Ed.: Karger-Kocsis J., Kluwer Academic, Dordrecht, 1999, pp Svensson N., Shishoo R., Gilchrist M., J. Thermoplast. Compos. Mater., 11 (1998), 22 5 Karger-Kocsis J., Czigány T., Mayer J., Plast. Rubb. Compos. Process. Appl., 25 (1996), Leong K.H., Ramakrishna S., Huang Z.M., Bibo G.A., Composites A, 31A (2000) Polymers & Polymer Composites, Vol. 9, No. 8, 2001

7 A Comparison of the Mechanical Behaviour of Weft-Knitted Glass and Carbon Fiber Fabric-Reinforced Polyamide-12 7 EMS hybrid yarns (product brochure), EMS Chemie, Domat, Switzerland, Mayer J., De Haan J., Kirch M., Reber R., Wintermantel E., Wild U., J. Thermoplast. Compos. Mater., 12 (1999), Czigány T., Karger-Kocsis J., Polym. Polym. Compos., 1 (1993) Karger-Kocsis J., Czigány T., Appl. Compos. Mater., 4 (1997), Jones D. P., Leach D. C., Moore D.R., Plast. Rubb. Compos. Process. Appl., 6 (1986) Denault J., Composite Interface, 2 (1994) Putnoki I., Moos E., Karger-Kocsis J., Plast. Rubb. Compos., 28 (1999), Karger-Kocsis J., Czigány T., Composites A, 29A (1998), 1319 Polymers & Polymer Composites, Vol. 9, No. 8,

8 T. Czigány and J. Karger-Kocsis 498 Polymers & Polymer Composites, Vol. 9, No. 8, 2001