ACKNOWLEDGMENTsuming that a crack is present and designing to an extension of the Paris Law at 4.2K. sample fatigue limit is reasonably safe, and even greater design safety can be ensured by as small-grained material are very high. Hence, a design that maintains loads below the smooth However, the values of Ao taht are required to raise AK to near AK[h for small cracks in a ber, and, certainly, is not ·a reliable design parameter value, when short cracks are present.The results suggest that the threshold stress intensity, AKG], is not a very meaningful num tigue endurance limit that can be determined at 4.2 K with ground and polished specimens. tion (2) is satished, a small crack does not grow when the applied stress is less than the fa tests at 4.2 K, and is bounded by tests at constant AKmax. If neither condidon (1) nor condi crack; its growth rate obeys the constitutive relations determined from the long-crack growth size or (2) the crack tip plastic zone is larger than grain size, the crack behaves as a long tion at 4.2 K. When either (1) the size of a crack is larger than several times the average gain 908, the experimental results suggest a method for estimating the rate of small crack propaga N0 matter what mechanisms control the anomalous growth of small crack in Incoloy elsewhere.10 OCR Output are summarized here; a complete description of our research on this subject will be published and (3) direct observations of the propagation of very small surface cracks. The main results long crack growth rates under closure-free conditions to simulate short crack propagation, tensity (Kmax), the method proposed by Herman, Hertzberg, and Jaccardg to determine the tests at constant load-ratio (R), (2) long-crack growth tests at constant maximum stress in We measured and compared fatigue crack growth rates in (1) conventional long-crack growthThe approach taken in this study resembles that in our previous work on A286 steel. isotropic continuum is no longer held. vant microstructural dimensions, the assumption that the crack grows in a homogeneous, fracture mechanics becomes invalid. (3) When the crack size is small compared to the rele local plasticity, the small-scale yield assumption that underlies the application of linear elastic when the crack is very short. (2) When the crack size becomes comparable to the scale of Crack closure that lowers the driving force for lon g-crack propagation is absent or reduced nisms for the anomalous growth of small cracks are described in recent reviews.5'7 (1) may propagate at cyclic stress intensities below the long-crack threshold. Several mecha cracks often grow faster than long cracks at the same nominal cyclic stress intensities, and tive relations that govem long cracks become invalid when the crack size is small; small Recent results on the propagation of surface fatigue cracks2'4 show that the constitu timating small crack propagation at 4.2 K. gation at room temperature and 77 K, and use the results to propose a feasible method for es tigue crack propagation tests at ...