Specimen designs for longitudinal tensile testing of unidirectional composites: an experimental and numerical study

The longitudinal tensile strength is one of the most basic characteristics of a unidirectional fibre-reinforced composite. Both ASTM and ISO have standardised this test, but they mostly result in failure near the gripped section. This failure mode is not considered an acceptable failure mode, as it leads to an underestimation of the actual strength value. There has been considerable research into mitigation strategies for this problem, including optimising the end tabs [1-3] or changing the specimen shape [4, 5]. While some studies combine experiments and models, they remain focused on one particular strategy or specimen design. This paper, however, focuses on a wide range of specimen designs, while still combining experiments and models. The initial testing phase involved the examination of five distinct specimen and tab designs. These comprised four rectangular specimens, each featuring a different end tab configuration: (1) straight end tabs, (2) continuous end tabs (also called sandwich specimens), (3) arrow-shape end tabs, and (4) circle-shape end tabs. As a fifth specimen design, we used the butterfly-shape specimen with tapered end tabs. All panels were manufactured using unidirectional HS40/736LT carbon/epoxy prepregs, provided by North Thin Ply Technology (Switzerland). The laminates were manually laid up in a [010] layup and cured in an autoclave. The average cured ply thickness was 0.5 mm. For the continuous tab specimens, four additional layers of S-glass/913 epoxy (Hexcel, UK) were added on the top and bottom, adding 1.2 mm of thickness in total. The end tabs were cut from 0/90 woven glass/epoxy composite panel with a nominal thickness of 2 mm. The tensile tests were performed on a Zwick Z100 tensile machine with hydraulic grips and a 100 kN load cell. The hydraulic pressure was adjusted to the lowest level that did not result in slippage. The surface strain distribution was monitored using full-field digital image correlation and processed using Vic-2D software. The results of the tensile test demonstrated that the inclusion of end tabs yielded considerable benefits. The rectangular specimens with rectangular end tabs, which are standardised by ISO and ASTM, exhibited a lower failure strain than the other three rectangular specimen designs. The circle and arrow-shape end tabs demonstrated performance comparable to that of the continuous tabs, although they did exhibit a slightly lower standard deviation. The butterfly design performed the worst in terms of average failure strain. We believe this is mainly due to the damage introduced by the water jet cutting. As the arrow-shape and circle-shape end tabs represent novel designs, we decided to investigate them in more detail with a finite element model. Both the arrow- and circle-shape tabs result in stress concentrations at the tab’s edge, albeit to a limited extent in magnitude and in a relatively small region. The circle-shape end tabs perform even more effectively, as they lead to even smaller stress concentrations than the arrow-shape end tabs. This helps to explain the higher failure strain they managed to achieve. Furthermore, the finite element model also helped to optimise the shape of the end tab.