Fatigue analysis of as-built and machined 308L steel specimens produced by wire and arc additive manufacturing

Additive manufacturing (AM) processes offer numerous advantages in the production of steel components, such as material efficiency and geometric flexibility. Such favourable features lead to an increasing application of these technologies in different working areas. However, there is still much to be learned about the fatigue behaviour of components produced using these techniques. The overall goal of this research is to investigate the fatigue strength of specimens obtained by the Wire Arc Additive Manufacturing (WAAM) technology. After some initial tests to characterize the behaviour of the material, one of the factors that contributes to the fatigue life of a component is investigated, that being the presence of residual stresses in the component. Another, even more critical factor in the fatigue strength of a WAAMed component is the effect of surface roughness. Surface defects such as spatters and deep notches, combined with the toe effect of welds, act as stress concentration points and can introduce cracks that lead to part failure. To analyse this effect, the fatigue resistance of WAAMed as-built and machined surfaces, as well as of the defective material is then investigated. The first step is to create a dedicated design geometry, based on the International Standard ASTM E466-21, for the specimens to be fatigue-tested. To evaluate the fatigue behaviour of the as-built surface and the machined base material, two types of specimens are used – respectively, one with 2 machined-surfaces and 2 as-built surface, and the other with all the surfaces machined – and two load ratios are tested, R 0.1 and R 0.5. From the failed samples, it is then possible to observe whether the fatigue failure initiation occurred from internal defects or from the as-built surface thanks to the SEM. With those results, the fatigue resistance curves for this material, in both as built and machined conditions, have been plotted. From the S-N curves it is also possible to experimentally derive the fatigue notch factor Kf, which can then be compared with a value of Kf estimated from some finite element analyses, carried out on the 3D scanning surface of the WAAMed plate (from which the specimens have been extracted) and following the FKM guidelines. This comparison allows to validate the method proposed in the FKM for these type of WAAMed components. Regarding the damage tolerance in the presence of defects, the Kitagawa-Takashi diagram is defined for a load ratio R=-1 by estimating the defect-free plain material fatigue limit and the threshold SIF of long cracks. To determine the fatigue limit asymptote, the main approach consists of a series of rotating bending tests on smooth specimens by employing the staircase method and the application of Murakami’s formula. For the determination of the threshold SIF range for long crack propagation, a preliminary estimation has been made thanks to a correlation suggested by Atzori-Meneghetti-Susmel and directives according to fracture mechanics.