a b s t r a c tThe plane-strain crack subjected to mode I cyclic loading under small scale yielding was analysed. The influence of the load range, load ratio and overload on the near-tip deformation-, stress-and strain-fields was studied. Although the near-tip zones of appreciable cyclic plastic flow for all loading regimes matched closely one another, when scaled with (DK/r Y ) 2 , the activities of plastic flow within them manifested dependence on K max and K min , as well as on overload. Cyclic trajectories of the crack-tip opening displacement (CTOD) converged to stable self-similar loops of the sizes proportional to DK 2 and positions in CTOD-K plane dependent on the maximum K along the whole loading route, including an overload. Computed near-tip deformation evidenced plastic crack advance, this way visualising of the Laird-Smith concept of fatigue cracking. This crack growth by blunting-resharpening accelerated with rising DK and was halted by an overload. Crack closure upon unloading had no place. The affinities were revealed between computed near-tip stress-strain variables and the experimental trends of the fatigue crack growth rate, such as its dependence on K max and K min (or DK and K max ), and retardation by overload. Thus, the effects of loading parameters on fatigue cracking, hitherto associated with crack closure, are attributable to the stress-strain fields in front of it as the direct drives of the key fatigue constituents -damage accumulation and bond breaking.

No abstract

No abstract

Continuum modelling of hydrogen diffusion in metals, which accounts for both trapping and an imposed force field, is revisited. A generalised model of hydrogen diffusion and trapping is developed as a continuous interpretation of the discrete random-walk theory. A system of nonlinear equations describing the phenomenon of diffusion with multiple types of traps is derived without the assumption of a local equilibrium among hydrogen populations in dissimilar positions. Lattice-trap interchange kinetics can degenerate into local equilibrium as a limit case. Moreover, certain terms in general equations may be negligible in specific situations. By removing these terms, known particularised models of hydrogen diffusion and trapping are recovered. Determining the terms, which are disregarded in reduced models, enables a straightforward assessment of the applicability of these models. The advantages and limitations of particularised models applied to hydrogen embrittlement analyses are discussed.

No abstract