M. Shioya, M. Nakatani/Composites Science and Technology 60(2000)219-229 and a fracture surface of the fibres characteristic to fibres. In the case of the axial compression bending test flexural fracture were observed near the transverse frac- of the composite strand, detection of the initial fracture ure surface of the composite strands, it was considered by using some method, such as the acoustic emission that the transverse fracture surface of the composite technique, is required in addition to detect fracture from trands resulted from microbuckling of the fibres. During the observation of the fracture process or from the drop the axial compression test of the PAN-based carbon of bending load in the load-displacement curve fibre composite strands with a low modulus matrix, the composite strand was deformed into a wavy shape without causing fractures of the fibres and matrix resin. References The difference of the fracture surface could be related to the fibre cross-section textures. The pitch-based carbon [1 Hawthorne HM fibres used in this study had a pleat-like texture extending carbon fibres. J M radially from the center of the cross-section, while the 2 Jones WR, Johnson Jw. Intrinsic strength and non-Hookean behaviour of carbon fibres. Carbon 1971- 9: 645-55 PAN-based carbon fibres had a more isotropic and 3 Allen SR. Tensile recoil measurement of compressive strength for homogeneous cross-section. eric high performance fibres. J Mater Sci 1987: 22: 853-9 The compressive strength of the PAN-based carbon 4 Jiang H, Abhiraman AS, Tsui K. Analysis of failure in ' recoil fibre composite strand increased with the matrix mod from tension'of PAN-based carbon fibres. Carbon 1993 31: 887- ulus. For the pitch-based carbon fibres, the matrix 5 Dobb MG, Guo H, Johnson DJ, Park CR. Structure-compres modulus dependence of the compressive strength of the sional property relations in carbon fibres. Carbon 1995: 33: 1553- composite strand was relatively small. The reduced compressive strength of carbon fibres determined using a (6 Fawaz SA, Palazotto AN Tensile and compressive properties of stiff matrix resin and the compressive strength determined rocl989:134:381-8 with the recoil test were almost in proportion to the com [7 Macturk Ks, Eby RK, Adams ww. Characterization of com- pressive strength determined with the micro-compression ressive properties of high-performance polymer fibres with a test. The ratio of the reduced compressive strength deter- microcompression apparatus Polymer 1991: 32: 1782-7 mined using a matrix with a tensile modulus of 3.0 GPa [8 Shinohara AH, Sato T, Saito F, Tomioka T, Arai Y. A novel against the compressive strength determined with the method for measuring direct compressive properties of carbon micro-compression test was 1. 15. The ratio of the com- fibres using a micro-mechanical compression tester. J Mater Sci 993:28:6611-6. pressive strength determined with the recoil test agains [9 Nakatani M, Shioya M, Yamashita J Axial compressive fracture that determined with the micro-compression test was of carbon fibres. Carbon 1999: 37: 601-8 0.74. The bending test on the composite strand gave [0 Ohsawa T Miwa M, Kawade M. Axial compressive strength of larger reduced compressive strength than the strength of arbon fibre. J Appl Polym Sci 1990: 39: 1733-43 [1 Shioya M, Nakatani M, Nakao K, Takaku A. Axial compression the fibres dtermined with other methods bending tests on carbon films and carbon fibre composites. J The -compression test gives more reliable value Mater Sci1999;:34:1301-11 of the compressive strength of bare fibres, while it 12] Testing method for tensile properties of plastics, Japanese indus requires laborious testing procedure. The recoil test is trial standard. 1981: JIS K7113 more easy to carry out and is a useful method for com- [3 Testing methods for carbon fiber content and void content of arbon fiber reinforced plastics, Japanese industrial standard, paring the relative compressive strength of a large 991;JISK7075 number of carbon fibres. It should be noted however, [14 Kozey VV, Jiang H, Mehta VR. Kumar S Compressive behavior that if flexural fracture occurs during the recoil test, the of materials: Part Il. High performance fibers. J Mater Res obtained value underestimates the compressive strength 95;10:10446 of the fibre. The axial compression test of the composite [5 Iwai H, Uemura M, Hayashi T. Issues of 3-point and 4-point strands is useful for obtaining the cor nodes in flexural testing methods for advanced composite mate- rials. J Jpn Soc Comp Mater 1992: 18: 60-5 of the fibres surrounded by the matrix, while the results [16] significantly depends on the stiffness of the matrix Thus, if the compressive strength of the component fibres are estimated using the axial compression test of constants of compression-annealed pyrolytic graphite. J Appl hys197041:3373-82. the composite strand, it is essential to properly select the [18] Rosen BW. Fibre composite materials. Materials Park, Ohio matrix resin especially in the case of PAN-based carbon American Society for Metals, 1965. p. 72-5.[Chapter 3].and a fracture surface of the ®bres characteristic to ¯exural fracture were observed near the transverse frac￾ture surface of the composite strands, it was considered that the transverse fracture surface of the composite strands resulted from microbuckling of the ®bres. During the axial compression test of the PAN-based carbon ®bre composite strands with a low modulus matrix, the composite strand was deformed into a wavy shape without causing fractures of the ®bres and matrix resin. The di€erence of the fracture surface could be related to the ®bre cross-section textures. The pitch-based carbon ®bres used in this study had a pleat-like texture extending radially from the center of the cross-section, while the PAN-based carbon ®bres had a more isotropic and homogeneous cross-section. The compressive strength of the PAN-based carbon ®bre composite strand increased with the matrix mod￾ulus. For the pitch-based carbon ®bres, the matrix modulus dependence of the compressive strength of the composite strand was relatively small. The reduced compressive strength of carbon ®bres determined using a sti€ matrix resin and the compressive strength determined with the recoil test were almost in proportion to the com￾pressive strength determined with the micro-compression test. The ratio of the reduced compressive strength deter￾mined using a matrix with a tensile modulus of 3.0 GPa against the compressive strength determined with the micro-compression test was 1.15. The ratio of the com￾pressive strength determined with the recoil test against that determined with the micro-compression test was 0.74. The bending test on the composite strand gave larger reduced compressive strength than the strength of the ®bres dtermined with other methods. The micro-compression test gives more reliable values of the compressive strength of bare ®bres, while it requires laborious testing procedure. The recoil test is more easy to carry out and is a useful method for com￾paring the relative compressive strength of a large number of carbon ®bres. It should be noted, however, that if ¯exural fracture occurs during the recoil test, the obtained value underestimates the compressive strength of the ®bre. The axial compression test of the composite strands is useful for obtaining the compressive strength of the ®bres surrounded by the matrix, while the results signi®cantly depends on the sti€ness of the matrix. Thus, if the compressive strength of the component ®bres are estimated using the axial compression test of the composite strand, it is essential to properly select the matrix resin, especially in the case of PAN-based carbon ®bres. In the case of the axial compression bending test of the composite strand, detection of the initial fracture by using some method, such as the acoustic emission technique, is required in addition to detect fracture from the observation of the fracture process or from the drop of bending load in the load-displacement curve. References [1] Hawthorne HM, Teghtsoonian E. Axial compression fracture in carbon ®bres. J Mater Sci 1975;10:41±51. [2] Jones WR, Johnson JW. Intrinsic strength and non-Hookean behaviour of carbon ®bres. Carbon 1971;9:645±55. [3] Allen SR. Tensile recoil measurement of compressive strength for polymeric high performance ®bres. J Mater Sci 1987;22:853±9. [4] Jiang H, Abhiraman AS, Tsui K. Analysis of failure in `recoil from tension' of PAN-based carbon ®bres. Carbon 1993;31:887± 94. [5] Dobb MG, Guo H, Johnson DJ, Park CR. Structure-compres￾sional property relations in carbon ®bres. Carbon 1995;33:1553± 9. [6] Fawaz SA, Palazotto AN. Tensile and compressive properties of poly(p-phenylene benzobisoxazole) ®ber. Mater Res Soc Symp Proc 1989;134:381±8. [7] Macturk KS, Eby RK, Adams WW. Characterization of com￾pressive properties of high-performance polymer ®bres with a new microcompression apparatus. Polymer 1991;32:1782±7. [8] Shinohara AH, Sato T, Saito F, Tomioka T, Arai Y. A novel method for measuring direct compressive properties of carbon ®bres using a micro-mechanical compression tester. J Mater Sci 1993;28:6611±6. [9] Nakatani M, Shioya M, Yamashita J. Axial compressive fracture of carbon ®bres. Carbon 1999;37:601±8. [10] Ohsawa T, Miwa M, Kawade M. Axial compressive strength of carbon ®bre. J Appl Polym Sci 1990;39:1733±43. [11] Shioya M, Nakatani M, Nakao K, Takaku A. Axial compression bending tests on carbon ®lms and carbon ®bre composites. J Mater Sci 1999;34:1301±11. [12] Testing method for tensile properties of plastics, Japanese indus￾trial standard, 1981; JIS K7113. [13] Testing methods for carbon ®ber content and void content of carbon ®ber reinforced plastics, Japanese industrial standard, 1991; JIS K7075. [14] Kozey VV, Jiang H, Mehta VR, Kumar S. Compressive behavior of materials: Part II. High performance ®bers. J Mater Res 1995;10:1044±61. [15] Iwai H, Uemura M, Hayashi T. Issues of 3-point and 4-point modes in ¯exural testing methods for advanced composite mate￾rials. J Jpn Soc Comp Mater 1992;18:60±5. [16] Fukuda H. Compression bending test method for advanced composites. J Jpn Soc Aeronautical and Space Sci 1993;41:482±7. [17] Blakslee OL, Proctor DG, Seldin EJ, Spence GB, Weng T. Elastic constants of compression-annealed pyrolytic graphite. J Appl Phys 1970;41:3373±82. [18] Rosen BW. Fibre composite materials. Materials Park, Ohio: American Society for Metals, 1965. p. 72±5. [Chapter 3]. M. Shioya, M. Nakatani / Composites Science and Technology 60 (2000) 219±229 229