Extreme value statistics and fracture mechanics for assessing the effect of as-built surfaces on the fatigue thresholds of additively manufactured light alloys and steels

The assessment of the effect of surface quality on the fatigue behaviour of additively manufactured parts is essential for guaranteeing the structural durability. This investigation applies a recently proposed fracture mechanics model based on an equivalent crack size definition that accounts for the weakening effect of as-built surfaces on the fatigue limit of additively manufactured parts. The purpose is to further validate the model, starting from the as-built surface texture characterisation by optical profilometry. More in detail, the equivalent crack size is associated to the deepest surface micro-notch of the as-built surface, which is estimated starting from optical profilometry data and subsequent statistical analysis of the distribution of micro-notches in individual batches of specimens through extreme value statistics. The model has been applied to unpublished constant amplitude fatigue test results of Scalmalloy® and additional experimental data taken from the literature and relevant to Al-, Ti-alloy, and steel specimens incorporating the effect of residual stresses. Theoretical estimates of fatigue thresholds resulted in good agreement with experimental values, where 75% of the dataset showed a percentage error lower than 20%, while the maximum deviation across the entire dataset was 26%.