Numerical reinterpretation of hydrogen permeation tests with residual stresses and consequences in crack modelling

Abstract

Hydrogen embrittlement involves different physical mechanisms around the crack tip, requiring a substantial modelling effort to predict critical conditions for hydrogen assisted cracking. However, the experimental characterisation of hydrogen traps still poses challenges in the interpretation of parameters, especially trap densities and energies. Electrochemical permeation is the most common method to obtain these parameters, and it has been extensively used to assess the relationship between plastic strain and trap density. However, even in cold rolled specimens the presence of residual stresses is usually neglected. This hypothesis is reconsidered in a finite element analysis in which a permeation test is reproduced, introducing the residual stress distributions as initial conditions. Due to the stress-drifted diffusion, the shape of these residual distributions produces a deviation in the analytical mapping strategy that is used to unequivocally determine trap densities and binding energies. The results indicate that the values fitted through an analytical mapping without considering residual stresses cannot be used to realistically reproduce the hydrogen concentrations near a crack tip.