Process-based fatigue qualification of printed metals

Two printed metal parts made to the same drawing can pass an initial strength test yet last different times under repeated loading. How can the printing conditions predict that spread in crack-initiation life? Melting powder with a laser leaves tiny voids, grain arrangements and residual stresses that affect where a fatigue crack first forms.[1]

Repeated loading can start a crack at a weak region of a printed metal part. Can its printing conditions predict when that happens across many parts, including those that fail much earlier than the average? The September 2026 preprint models fatigue initiation using effective crack-path energy, elastic-energy release and stored plastic work per cycle. These quantities are correlated through the same microstructure. Grain maps and X-ray images of voids inform its printed-steel models; analytical distributions agree with numerical evaluations. The example correlations are prescribed, not inferred from a population of fatigue failures.[1] [2]

Models already predict lifetime scatter from measured defects in completed parts. The missing step is a validated prediction starting from printing conditions and carrying uncertainty through the resulting microstructure to fatigue risk. A calculation that assumes the necessary correlations has not yet established this manufacturing-based qualification.[1] [2]

The authors explicitly retain experimental calibration of those correlations and connection to microstructure prediction from print parameters. This boundary matters because May 2026 work already reproduced fatigue-life scatter and failure-mode changes using measured surface and subsurface defect statistics in maraging steel. The unresolved target starts earlier in the process chain and requires its uncertainty to remain valid when manufacturing conditions change.[1] [2] [3]

Changing a printing setting can change the internal material as well as the part’s shape. A reliable process-based forecast would help establish what fatigue evidence transfers after that change. It would make the reasoning behind qualification explicit, including where additional experiments are still needed.[1] [3]

A part that survives an initial load can still fail after repeated loading, so qualification needs evidence about early failures and scatter as well as strength or mean life. Linking printing conditions to those risks would explain when a process change justifies reusing earlier fatigue evidence and when new tests are needed. Both the process-to-material and material-to-failure links must be validated on specified alloys, geometries and loads before the prediction can support that reasoning.[1] [3]