Szczegóły publikacji
Opis bibliograficzny
Probabilistic prediction of fatigue life scatter from surface and subsurface defect populations in additively manufactured metals / Aleksander Karolczuk, Marta Kurek, Mihiretu Ganta, Aleksander Hebda, Piotr SKUBISZ, Małgorzata WITKOWSKA // Additive Manufacturing ; ISSN 2214-8604 . — 2026 — vol. 124 art. no. 105236, s. 1–19. — Bibliogr. s. 18–19, Abstr. — Publikacja dostępna online od: 2026-05-08
Autorzy (6)
- Karolczuk Aleksander
- Kurek Marta
- Ganta Mihiretu
- Hebda Aleksander
- AGHSkubisz Piotr
- AGHWitkowska Małgorzata
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 167816 |
|---|---|
| Data dodania do BaDAP | 2026-05-22 |
| Tekst źródłowy | URL |
| DOI | 10.1016/j.addma.2026.105236 |
| Rok publikacji | 2026 |
| Typ publikacji | artykuł w czasopiśmie |
| Otwarty dostęp |  |
| Czasopismo/seria | Additive Manufacturing |
Abstract
Fatigue life scatter in additively manufactured (AM) metals is controlled by competing crack initiation mechanisms associated with surface and subsurface defects. This work presents a probabilistic fatigue framework that predicts fatigue life distributions directly from measured defect populations, without introducing empirical life-scatter terms or predefined fatigue-active volumes. Surface and subsurface defects are modeled as independent Poisson point processes, and the upper tails of their size distributions are described using Generalized Pareto distributions fitted to X-ray computed tomography and optical profilometry data. A Murakami-based defect relation, including an effective stress-ratio correction for retained near-surface residual stresses, links defect size and local stress amplitude to fatigue life. A defect-size-dependent admissible volume is introduced to ensure geometric consistency under non-uniform stress fields. The framework is validated using plane-bending fatigue data for additively manufactured maraging steel fabricated at different build orientations. The results demonstrate that stress-dependent fatigue life scatter, quantile separation, and transitions between surface- and subsurface-controlled failure modes can be reproduced using a single material model driven primarily by defect population statistics. The framework also predicts size-dependent fatigue behavior and transitions between competing crack initiation mechanisms without introducing additional size-scaling parameters.
