Abstract
The fiber straightening pre‐tension applied during the sample preparation of single fiber composites is suggested here to significantly affect the number of fragments as well as the value of the Kelly‐Tyson interface shear strength obtained from single filament composite tests. This implies that fragmentation tests performed under seemingly identical conditions, but in which the fiber pre‐tension is not accurately controlled during sample preparation, might yield widely differing experimental results. We present and discuss a new set of experimental results dealing with fiber pre‐tension effects in carbon/epoxy single fiber composites.