PSI - Issue 76

Luca Esposito et al. / Procedia Structural Integrity 76 (2026) 50–58

57

Fig. 3. Left: 3D probability density function ( p df ) surface showing the evolution of fatigue life distribution as a function of stress amplitude and number of cycles. Right: p df cross-sections at S = 200MPa and S = 300MPa.

were triggered by porosity. This alternative model provides a significantly better fit and more realistic representation of the fatigue life scatter for this dataset.

4. Conclusions

This study presents a defect-sensitive probabilistic framework for modeling the fatigue behavior of AlSi10Mg specimens produced via Selective Laser Melting (SLM), with particular focus on the pronounced life scatter observed in Z-oriented, machined samples. Single-mode Weibull formulations, though adequate for as-built samples, were shown to be insu ffi cient for describing the variability in machined specimens, where internal defect morphology becomes the dominant factor in fatigue performance. A novel bimodal log-Weibull model was developed and calibrated using Maximum Likelihood Estimation (MLEM), enabling the separation of fatigue life distributions associated with two distinct defect types: porosity and lack-of-fusion. The model was validated through fractographic analysis, which confirmed the coexistence of these two failure mechanisms and supported the interpretation of lifetime scatter as a consequence of defect-type competition. The bimodal formulation captures both the median trend and the observed variability in fatigue life more accurately than conventional approaches. In particular, it reproduces the transition from narrow, single-peaked distributions at high stress levels to broader, bimodal distributions at lower stresses, in line with experimental evidence. The frame work also allows for the estimation of survival probabilities across stress levels, improving the predictive capability of fatigue life assessment under realistic conditions. Overall, the proposed model provides a physically grounded and statistically robust approach to fatigue life pre diction in additively manufactured materials. By explicitly incorporating the role of internal defects, it supports the development of more reliable design criteria and enhances the understanding of fatigue failure mechanisms in AM components.

References

ASTM-E739, 2015. Standard Practice for Statistical Analysis of Linear or Linearized Stress-Life (S-N) and Strain-Life ( ϵ -N) Fatigue Data. ASTM International , 1–7doi: 10.1520/E0739-10R15 . Brandl, E., Heckenberger, U., Holzinger, V., Buchbinder, D., 2012a. Additive manufactured alsi10mg samples using selective laser melting (slm): Microstructure and mechanical properties. Materials & Design 34, 159–169. Brandl, E., Heckenberger, U., Holzinger, V., Buchbinder, D., 2012b. Additive manufactured alsi10mg samples using selective laser melting (slm): Microstructure, high cycle fatigue, and fracture behavior. Materials & Design 34, 159–169.