Abstract
As well known, a non-propagating fatigue crack is easy to be formed at the root of a notch as the radius of curvature of the notch decreases. Although a crack may be considered to be an extremely sharp notch with an infinitesimal radius, it is not clear yet whether non-propagating cracks are developed at the tips of the cracks or not. To clarify this, some experiments were made on two kinds of specimens, the plain specimen with a crack and the notched specimen with a crack. It was found that a non-propagating crack could not be developed at the tip of the crack in the former specimen, but could be in the latter specimen.
For the interpretation of these results, the experimental facts found by Frost are referred to, which show that crack initiation is governed by the alternating stress amplitude (independent of mean stress), and crack propagation is determined by the maximum tensile stress in the cycle. Since the crack existing in the plain specimen does not bear the compressive stress concentration, the alternating stress at the crack tip will be relatively low. On the other hand, the alternating stress will be larger at the tip of the crack existing at the notch root, because of the stress field associated with the notch, operating even in the compression phase. From such a difference in stress conditions, coupled with cyclic hardening, satisfactory explanation for the experimental results can be derived.
