Several mechanisms have been proposed in the literature to account for crack growth retardation following a single peak overload.
In order to determine the dominant mechanism, overload tests were performed on thick and thin specimens made from BS4360 50B structural steel. The baseline stress intensity range was 25 MPa √m and the load ratio (= minimum load/maximum load of fatigue cycle) was 0.05, while the overload was of stress intensity range 50 MPa √m.
It was observed that the crack growth and closure responses were different at the surface and in the bulk of the thick specimen; no such variations in behavior occurred along the crack front for the thin specimen. For both thicknesses of material, the crack growth rate predicted by measurements of the crack opening load was in agreement with the observed crack growth rate, except for the period when crack growth rates were recovering from the slowest transient growth rate to the post-overload stabilized value. This discrepancy was due to the phenomenon of discontinuous closure—the crack first closed at a location far from its tip. Fracture surface profiles showed that the crack path deviated by only a small amount after application of the overload; hence the retarded growth cannot be due to crack branching.
It is concluded from these tests and from a critical examination of the literature that, at high baseline ΔK levels, retardation is due to plasticity-induced crack closure. At low baseline ΔK levels approaching the threshold value, retardation may be due to plasticity-induced crack closure or to irregularities of the crack front.