According to the hydrogen induced decohesion (HID) theory, hydrogen causes fracture by weakening the bonds in a steel lattice, such that they eventually break at lower applied stresses than would be requisite in air. Since hydrogen-induced fracture occurs rapidly, hydrogen surrounding the “free surfaces being formed” is unlikely to have time to reach equilibrium with the surrounding “stress and strain fields” , and the concentration at the tip of a growing crack is thus likely to be much greater than it would be at equilibrium.
Moreover, crack fronts are theorized to have regions in which non-Hookean elastic stresses become large enough that they significantly lower the dissolved hydrogen’s chemical potential. This lower potential drives additional diffusion, such that the hydrogen concentrations that develop in these regions are ultimately several orders of magnitude greater than they would be at equilibrium 43. The HID theory was once thought to be supported by experimentation showing that a stationary crack in a wedge open loaded (WOL) specimen being firmly held in place in a hydrogen chamber can be induced to propagate solely by increasing the hydrogen pressure; it was originally thought that such propagation could not occur due to any mechanism that depends on dislocation mobility, but the validity of this claim has since been argued 34. This experimentation moreover demonstrated the existence of a clear inverse relationship between the threshold hydrogen pressure at which a stationary crack will propagate and the stress intensity being applied to the crack; this relationship suggests that, above a given threshold, hydrogen reduces the cohesive strength between the atomic bonds in a steel lattice enough that they break. HID is also supported by analyses showing that fracture in the presence of hydrogen frequently occurs without significant local deformation 34, 44.”