Long-term thermo-oxidative aging in composite materials: Experimental methods

The combination of this work, an earlier study [1] concerning the effects of thermo-oxidative aging on the thermal, fracture and compressive behavior of carbon-fiber reinforced epoxy composites, and a companion paper [2] discussing the failure mechanisms due to thermo-oxidative aging provide the first published comprehensive study of the effects of long-term thermo-oxidative aging on such a wide range of properties for the same materials systems. In this phase of the work, tensile tests (both unidirectional 0 degrees and +/-45 degrees layups), edge delamination, and compression-after-impact (CAI) were examined for G30-500/R922-1 and G30-500/R6376 carbon/epoxy specimens aged in air at 177 degrees C (350 degrees F) up to 10,000 hours. Additionally, 10,000 hour data not included in the previous study are also included for comparison. The purpose of the present study is not to identify suitable candidate materials, but to evaluate the effect of test methods and failure modes in characterizing thermo-oxidative stability. The materials systems chosen were selected as model materials for the study of degradation phenomena and not for performance evaluation at 177 degrees C which is clearly beyond the useful range of these materials. Nevertheless, uni-directional tensile properties were found to be relatively unaffected up to 5,000 hours of aging. By contrast +/-45 degrees tensile strengths and unidirectional compression strengths decreased between 10 and 35% after only 1,000 hours exposure. Mode I fracture toughness increased with aging due to increased fiber bridging and Mode II fracture toughness was, by contrast, relatively unaffected. Edge delamination fracture toughness was severely affected by thermo-oxidative aging decreasing to between 5 and 15% of the unaged value after 10,000 hours aging suggesting a strong effect of interfacial degradation. Because even small weight losses lead to correspondingly large reductions in certain mechanical properties, even small errors in the determination of weight loss lead to large errors in estimates of mechanical properties. Indeed, it is likely that weight loss, even with the presence of corresponding mechanical data is not a good indicator of property degradation. Weight loss is only useful as a negative indicator: it indicates an unacceptable material, it does not yield useful information as to useful lifetimes.