ESTIMATION OF TENSILE STRENGTH DISTRIBUTION FOR CARBON FIBER WITH DIAMETER VARIATION ALONG FIBER

ESTIMATION OF TENSILE STRENGTH DISTRIBUTION FOR CARBON FIBER WITH DIAMETER VARIATION ALONG FIBER
复制标题

估算沿纤维直径变化的碳纤维拉伸强度分布

DOI:
10.2472/jsms.48.6appendix_90
复制
发表时间:
1998
期刊:
Journal of The Society of Materials Science, Japan
影响因子:
--
通讯作者:
N. Horikawa
N. Horikawa
中科院分区:
--
文献类型:
--
作者:
Tsuneshichi Tanaka;H. Nakayama;A. Sakaida;N. Horikawa

文献摘要

被引文献

相似文献

研究了不同标距长度(2 ~ 100 mm)碳纤维的拉伸强度分布。在拉伸试验中,通过使用光学显微镜,将纤维直径确定为在量规截面的两端和中心附近的三点最小直径d3。对数据和碳纤维拉伸强度的有效体积模型进行威布尔分析表明,不同标距长度的碳纤维形状参数a并不恒定,有效体积模型不能有效地应用于计算结果。发现这些是由于直径沿碳纤维从来不是沿着恒定的事实造成的。因此,在本研究中,旨在推导出真实最小直径d 0沿着标距的拉伸强度σ0的分布。为了做到这一点,d3和d 0的二维分布是从沿单根纤维每沿着0.1mm测量的直径数据导出的,所述单根纤维的标距长度为100 mm。将这些d3和d 0的分布与直径d3的抗拉强度σ3的分布相结合,解析地导出了强度σ0的分布,并利用所包含的参数的数值进行了计算。得到的各标距的σ0分布相互吻合得很好,可以用单一的正态分布来表示,这表明了本文分析的有效性。
Tensile strength distribution of carbon fibers was investigated for different gauge lengths ranging from 2 to 100mm. In tensile tests, the fiber diameter was determined as the three-point minimum diameter d3 at neighborhoods of both ends and the center of gauge section by using an optical microscope. Weibull analysis of the data and the effective volume model for tensile strength of carbon fiber showed that the shape parameter a was not constant for different gauge lengths and the effective volume model could not be applied effectively to the results. It was found that these were caused by the fact that the diameter was never constant along a carbon fiber. Consequently, in this study, it was aimed to derive the distribution of the tensile strength σ0 for the true minimum diameter d0 along the gauge length. In order to do this, two-dimensional distributions of d3 and d0 were derived from the data of diameter measured at every 0.1mm along a single fiber with the gauge length of 100mm. Combining these distributions of d3 and d0 with the distribution of the tensile strength σ3 for the diameter d3, the distribution of the strength σ0 was derived analytically, and calculated by using numerical values of included parameters. The distributions of σ0 obtained for respective gauge lengths mutually agreed very well and could be represented by a single normal distribution, and this showed the validity of the present analysis.