Failure in the junction region of T-stiffeners: 3D-braided vs. 2D tape laminate stiffeners

Strain distributions and failure mechanisms are compared for a three-dimensionally (3D) braided T-stiffener (preform designed and supplied by 3TEX Inc.) and a conventional two-dimensional (2D) tape laminate T-stiffener, bonded onto a tape laminate skin. The strain distributions in a pull-off test are measured by laser speckle interferometry and calculated by computational simulations. With good agreement between experiment and theory, substantial differences are found between the two classes of stiffener. The tape laminate stiffeners exhibit large strain concentrations across the noodle region and in the adjacent radii, which correlate well with observed first cracking events. The 3D-braided stiffeners show relatively uniform strain distributions throughout the flanges, the web, and the flange/web junction region. Strain concentrations are modest at the corner of the junction and absent along the interface between the flanges and the skin. Failure in the 3D-braided stiffeners does not occur within the junction region, but by a sequence of cracking events, first next to the junction region and then at the end of one flange. The pull-off load at the first failure event is substantially higher for the 3D-braided stiffener than the tape laminate stiffener, which is attributed mainly to the relative absence of strain concentrations in the former. In predicting strain distributions, account is taken of the details of the 3D architecture of the 3D-braided stiffeners as specified by the supplier (3TEX Inc.) by using the Binary Model of textile composites, whose formulation has been described previously. Comparison of the predicted and measured spatial distributions of strain constitutes a critical test of the Binary Model. It is found to perform well in this case.