A Combined Approach Using Critical Distance Theory and Extreme Value Statistics to Assess the Impact of Surface Roughness and Pores on the Fatigue Strength of Inconel 718

The aim of this study is to develop a predictive model to evaluate the impact of surface roughness and porosity on the fatigue strength of Inconel 718 specimens, both plain and V-notched, manufactured via laser powder bed fusion and tested in as-built and machined surface conditions. Fractographic analyses, combined with Gumbel and exponential probability distributions, were used to model pore diameters and their distances from the external surface. Surface morphology was characterized using a 3D optical profilometer. The identified probability distributions regarding the characterization of the pores were used to generate random values of pore diameters and distance from the external surface that were used in FE simulations as well as the surface profiles obtained through the optical profilometer. Once FE simulations were performed, critical distances, determined by analyzing machined blunt and sharp V-notched specimens, were combined with the Gumbel or generalized extreme value distributions to estimate the fatigue strength concentration factors due to pores and surface roughness at a 99% confidence level. After, the fatigue stress concentration factors of the pore and the surface roughness were combined to predict the fatigue strength of blunt V-notched specimens in the as-built state, as well as plain specimens in both as-built and machined conditions. The approach was implemented for specimens manufactured with different type of powder, i.e. the virgin, the cyclone and the recycled, to predict the fatigue strength and the number of cycles to failure. Finally, the comparison between predicted and experimental data is critically discussed.