Revisiting Classical Issues of Fatigue Crack Growth Using a Non-Linear Approach

Fatigue crack growth (FCG) has been studied for decades; however, several aspects are still objects of controversy. The objective here is to discuss different issues, using a numerical approach based on crack tip plastic strain, assuming that FCG is driven by crack tip deformation. ΔK was found to control cyclic plastic deformation at the crack tip, while K(max) has no effect. Therefore, alternative mechanisms are required to justify models based on ΔK and K(max). The analysis of crack tip plastic deformation also showed that there is crack tip damage below crack closure. Therefore, the definition of an effective load range ΔK(eff) = K(max) − K(open) is not correct, because the portion of load range below opening also contributes to FCG. Below crack closure, damage occurs during unloading while during loading the crack tip deformation is elastic. However, if the maximum load is decreased below the elastic limit, which corresponds to the transition between elastic and elasto–plastic regimes, there is no crack tip damage. Additionally, a significant effect of the crack ligament on crack closure was found in tests with different crack lengths and the same ΔK. Finally, the analysis of FCG after an overload with and without contact of crack flanks showed that the typical variation of da/dN observed is linked to crack closure variations, while the residual stresses ahead of crack tip are not affected by the contact of crack flanks.