A REVIEW OF CRACK CLOSURE, FATIGUE-CRACK THRESHOLD AND RELATED PHENOMENA

Published fatigue crack closure mechanisms are reviewed in the context of the following five ideas that have been developed over the past twenty years to explain the closure effects on near-threshold crack growth behaviour: (1) oxide, (2) asperity, (3) plasticity, (4) phase transformation and (5) viscous fluids. The first three have been considered as more important than the last two. Our analysis indicates that (a) there can be no contribution from plasticity to crack closure, (b) the crack closure contribution can be significant only if it is closed fully, which is reflected as an infinite slope in the load-displacement curve, (c) formation of oxide asperities from a fretting action is a random process and not a deterministic one, and therefore cannot explain the deterministic behaviour of the effect of load ratio on the threshold, and (d) the closure contribution from the asperities resulting from oxides or corrosion products, or surface roughness, it is less than 20% of what has been deduced based on the change in the slope of the load-displacement curves. Thus the analyses show that crack closure can exist, but its magnitude is either small or negligible. The critical evaluation of the literature data on (1) the threshold stress intensity variation with load ratio on many materials, and (2) and examination of the experimentally observed load-displacement curves confirm the above conclusions. Hence to rationalize the observed variations in the fatigue crack threshold with load ratio, we have proposed a new model. It postulates a requirement for two critical stress intensity parameters, namely Delta K-th* and K-max*, that must be satisfied simultaneously as the crack tip driving forces for fatigue crack growth. This requirement is fundamental to fatigue, since an unambiguous description of cyclic loads requires two independent load parameters. Several experimental results from the literature are presented in support of this postulation. Using these two critical parameters, the entire functional relationship between K-max, Delta K-th, and R is explained without invoking an extrinsic factor, namely crack closure. In addition, for a given material and its crack tip environment, a unique relationship between Delta K-th and K-max exists that is independent of test methods used in determining thresholds. Finally, because of this two parameter requirement, all fatigue crack growth data need to be represented in terms of three-dimensional plots involving da/dN, Delta K and K-max. For a two-dimensional representation, the data need to be transformed correctly, defining the net driving force involving both Delta K and K-max parameters. The concepts presented are independent of whether crack closure exists or not, or even whether cracks exist or not.