Effects of fiber gripping methods on the single fiber tensile test: I. Non-parametric statistical analysis

Single fiber tensile tests using two different gripping methods were carried out on various fiber lengths. One method (the glue-tab grip method) consists of mounting a fiber onto a rigid tab as specified in ASTM C1557-03 using an adhesive, while the other (the direct grip method) involves directly clamping a fiber using poly(methyl methacrylate) blocks. The tensile moduli obtained by the glue-tab grip as a function of the fiber gauge length are clearly different from those of the direct grip for fiber length between 2 and 10 mm, based on graphical statistical analyses using kernel density and q–q plots. This difference is caused by the gripping effect. In addition, the tensile strains and strengths obtained by the glue-tab grip were different from those of the direct grip at the short fiber gauge lengths (i.e., 2–10 mm). The differences between the tensile properties (i.e., modulus, strain, and strength) for the two grips measured with the 60-mm fiber length were not statistically significant compared to the results with the short fiber lengths.