Entropy-based fatigue characterization of metal additive manufacturing: a study on SLM 316L stainless steel

Abstract This study investigates the fatigue behavior of additively manufactured (AM) 316L stainless steel produced via selective laser melting (SLM) within an entropy-generation framework for fatigue evaluation. Flat dog-bone specimens with thicknesses of 2 mm and 3 mm were fabricated using optimized SLM process parameters, determined based on monotonic mechanical properties, and subsequently tested under cyclic loading at various stress amplitudes and frequencies. The fatigue performance of AM specimens was compared with that of conventionally manufactured (CM) 316L stainless steel to provide a comparative assessment of fatigue behavior influenced by manufacturing-induced features. Surface temperature evolution during cyclic loading was monitored using infrared thermography and analyzed through multiple thermodynamic approaches—including the self-heating rate (R θ ), cooling rate (R c ), hysteresis loop analysis, and a two-dimensional inverse heat conduction procedure—to compute the fracture fatigue entropy (FFE). The results reveal that AM specimens exhibit lower fatigue resistance and shorter fatigue lives than their CM counterparts, mainly due to the presence of porosity, lack of fusion, and residual-stress-induced anisotropy, which promote early crack initiation and multi-site failure. Despite differences in loading conditions, the FFE for AM specimens remained nearly constant, within 1.0–1.6 MJ/(m 3 K), confirming that fatigue failure occurs at a critical cumulative entropy level independent of frequency or R-ratio. In contrast, CM 316L specimens exhibited significantly higher FFE values—ranging from 11.3 to 13.5 MJ/(m 3 K) with an average of ≈ 13 MJ/(m 3 K)—reflecting their superior microstructural integrity. These findings highlight that the additive manufacturing process must be carefully controlled to avoid poor fatigue properties.