Some materials such as 7050-T7451 and β-annealed Ti–6Al–4V exhibit very rough fatigue crack surfaces which suggests that roughness induced crack closure might be a significant issue. The present work examined a range of data for 7050-T7451 material which included experimental results under constant amplitude and variable amplitude/spectrum loading. Crack closure was either measured directly, or the effects were inferred through results from previous fractographic investigations. Analyses using an updated and improved version of the FASTRAN crack growth code combined with the experimental data were used to identify a new approach to correlate closure levels and crack growth behaviour. The new approach includes direct compliance based crack opening measurements to correlate the baseline constant amplitude properties and then using a cycle-by-cycle strip yield model with realistic constraint factors to model crack growth under spectrum loading. The new approach is consistent with fundamental differences in fatigue crack closure levels under constant and variable amplitude/spectrum loading which have been identified at the individual cycle by cycle level down to very small crack sizes. The new approach is shown to work very well for a range of cases considered here, and it provides a very useful insight into this complex behaviour.
Read full abstract