The fatigue strength curves for fiber reinforced laminates are normally available at certain combinations of load ranges and minimum-to-maximum load ratios. The data are alternatively presented in the form of a constant life diagram. Real-world signals contain, however, a finite number of load ranges and mean values pairs which rarely coincide with those experimentally measured. For fatigue assessment purposes, an interpolation in the constant life diagram, seeking for fatigue strength curves not originally measured, is needed. It is the aim of this paper to present and apply a new way for interpolating in the constant life diagram. The interpolated fatigue strength curve is then used in a simulated fatigue life prediction case study. The source for fatigue loads comes from two normalized load signals of variable amplitude that can be scaled to any desirable (and reasonable) maximum stress maximum stress value. In this way, two predicted maximum stress versus cycles to failure can be computed. The relative position of these two curves agrees with the severity of each load signal which confirms the consistency of the proposed algorithm. Furthermore, the variable amplitude curves fall below the experimental ones (at a load ratio [Formula: see text]), that is, disregarding the interactions between large and small cycles implies in unacceptable (non-conservative) predictions in fatigue design.
Read full abstract