Fiber-reinforced composites, because of their superior specific strengths and stiffnesses, are used in many aircraft components. However, in this application these composites are subjected not only to fatigue loading, but to occasionally high velocity impact due to the bird injection, hail, dust, and rain. Thus, it is important to evaluate the residual life and degradation due to combined fatigue and impact loadings. Unidirectional graphite epoxy composites (MA8276-Tiger) which are used in the aerospace industry were impacted by a free falling weight at energy levels of 0.567j, 1.134j, and 1.571j [impact energy toughness (j/cm3); 0.12, 0.24, 0.34], respectively. The subsequent changes/degradation in elastic moduli, strength, toughness, and fatigue properties were measured after different number of impacts. It was found that for all energy levels these properties vary linearly with the number of impacts. Furthermore, attenuation changes is not a good ultrasonic parameter for degradation estimation, since it does not incorporate the micro- and macrocracks beyond the impact point. However, these micro- and macrocracks have significant effect on the mechanical properties. In contrast to the attenuation, the stress wave factor, which indicates the efficiency of wave propagation along the specimen, correlates very well with degradation, and it can be used effectively to measure the residual strength after impact. Ultrasonic characteristic on specimens subjected to combined fatigue and impact were also studied. Based on these experiments, it is concluded that the loss in fatigue residual life due to impact loads may be predicted by measuring the effects of the impact load on attenuation and stress wave factor. It was found that the reduction in fatigue life is proportional to sudden changes in attenuation and stress wave factor. Damage accumulation models based on Coffin-Manson equation, was suggested for impact and combined fatigue and impact. It was found that residual properties and fatigue life can be estimated from these models.
Read full abstract