We have re-evaluated short crack growth behavior using concepts developed recently, and we show that these concepts provide a unified framework that can explain both short and long crack growth behavior without resorting to the crack closure effect. We consider that the behavior of long cracks, including the effects of load ratio, R, is fundamental. We had shown previously that, since fatigue is at least a two-parameter problem in that at least two load parameters are required for an unambiguous description, there are two critical driving forces required simultaneously for fatigue cracks to grow. In extending this analysis to the growth of short cracks, we reject the current notion of the lack of similitude for short cracks and express the similitude as a fundamental postulate that, for a given crack growth mechanism, equal crack tip driving forces result in equal crack growth rates.
We find that, in most short crack problems considered in the literature, the cracks grow in pre-existing stress fields, which are either externally introduced or in situ generated. We refer to them as internal stresses. These can be residual stresses, residual stresses from prior treatment, or stresses generated during cycling. We show here that the observed lack of similitude and the large spread in the data for short cracks are due to: (1) an incomplete description of fatigue crack growth, (2) the variability of the internal stresses, and (3) uncertainties in their quantitative description (elastic versus elastic-plastic). We consider that the short cracks also grow at the same two thresholds as the long cracks, when all the components that contribute to the crack tip driving forces are included. In fact, short crack growth behavior confirms our concept that two parameters are required to define fatigue; consequently, for fatigue cracks to grow, two thresholds need to be satisfied simultaneously. We present examples from the literature to illustrate the concepts discussed.