IMPROVED ANALYSIS AND EXPERIMENTAL EVALUATION OF THE SINGLE FILAMENT PULL-OUT TEST

The single filament pull-out test has received attention as a preliminary test for screening fiber surface treatments that are designed to modify the interface-sensitive behavior of fiber composite materials. Its merits are that it is a direct test of adhesive bonding at the fiber/matrix interface, that it requires only a small amount of material, and that it can be used with either brittle or ductile matrix materials. In spite of its merits, the single filament pull-out test is still controversial. This is because the analysis of the test has not been proved indisputably to be correct, leaving the impression that the test is still not totally understood. In the work described in this paper, we improved the analysis of the single filament pull-out test by adding a matrix compression contribution to the stored strain energy and by including the work of friction between the debonded fiber and matrix in the energy balance. In addition, we were able to extend the experimental data over a much wider range of embedded length by using a strong fiber combined with a weak interface. This allowed a comprehensive comparison of theory with experiment, and showed that the improved model not only predicts the experimental data better than before, but also predicts the stable crack propagation observed experimentally for specimens with large embedded lengths. Methods for determining interfacial fracture energy from single filament pull-out data are also described.