Predicting fatigue life of pre-corroded 2024-T3 aluminum

Abstract

Much research has examined the viability of total fatigue life prediction methodologies based on the initial condition of the material. Further, corrosion attack is also of interest in terms of its damaging effects on structural durability. As a natural next step in understanding the effects of corrosion damage, this investigation was aimed at assessing the viability of a total fatigue life prediction methodology for material with pre-existing corrosion damage. The work covered here was mainly performed to further the breadth and scope of research and development, provide insights into several factors influencing fatigue resistance of corroded material with a basis in analytical and experimental data, and push closer to a fully conclusive analysis tool for truly predicting fatigue life in corroded materials. While the qualitative effects of corrosion and its effects on fatigue resistance are acknowledged by many researchers, a rigorous analytical technique for fully capturing the quantitative consequences of corrosion does not exist in closed form. Most current strategies call for overly conservative and costly repairs or make use of arbitrary factors of safety. For this study, fatigue specimens of nominal gage 0.063 in. aluminum alloy 2024-T3 were exposed to corrosion and tested in a laboratory setting. The test samples were taken from a single lot of material and corroded for 6, 24, and 72 h in three specimen orientations. The corroded specimens were cycled to failure at three stress levels. All failed specimens were examined to characterize the damage state(s) and failure mechanism(s). An analysis technique for predicting fatigue life results in the presence of corrosion damage was developed using a fracture mechanics foundation and the observed mechanisms of failure. In general, fatigue tests showed a decrease in life due to increased corrosion exposure. However, the fatigue tests from different specimen orientations exhibited similar lives to failure. The spread from shortest to longest fatigue lives among the different corrosion conditions decreased at the higher stress levels. Life predictions based on measured nucleation sites were generally conservative in nature and within 20–30% of the experimental lives. The main conclusion from this work was that a total fatigue life prediction methodology was shown to be a mostly successful tool in predicting the life of pre-corroded test specimens, while capturing the effects of three stress levels, three corrosion levels, and three specimen orientations.