A COMPARISON OF FAILURE CRITERIA TO PREDICT DELAMINATION OF UNIDIRECTIONAL GLASS EPOXY SPECIMENS WAISTED THROUGH THE THICKNESS

Both stress- and fracture-mechanics-based failure criteria for delamination are compared with experimental results of unidirectional laminates waisted through the thickness and loaded in tension. The fracture mechanics criterion predicts a reduction in strength with size which is not found experimentally. This indicates that the fracture mechanics criterion is not appropriate to predict delamination failure of the uncracked waisted specimens in which no macroscopic singularity exists. The maximum stress criterion is inaccurate because it does not account for the interaction between stress components. All the interactive failure criteria considered predict the delamination failure stress reasonably well because they are essentially curve fitting in nature. The Tsai-Wu criterion is of the most general form and gives a better prediction than the Tsai-Hill and Hashin-Rotem criteria. Any value within the range of -0.5 less-than-or-equal-to F*12 less-than-or-equal-to 0.9 for the normalized interaction term F*12 required in the Tsai-Wu quadratic criterion seems to be able to provide a prediction within the current engineering requirements. This range is wider than the commonly used range of -0.5 less-than-or-equal-to F*12 less-than-or-equal-to 0 in the literature.