Six-Stress-Ratio Fatigue Dataset for Mean-Stress Extrapolation in Laser Powder Bed-Fused 18Ni300 Steel
Description
This dataset accompanies the article: Karolczuk A. Role of mean-stress transformations in Gaussian process-based fatigue-life extrapolation of laser powder bed-fused 18Ni300 steel. International Journal of Fatigue 2026;213:109874. https://doi.org/10.1016/j.ijfatigue.2026.109874. The Excel file contains fatigue-test data for unmachined laser powder bed-fused 18Ni300 maraging steel tested under six stress ratios: R = −2, −1, −0.5, −0.25, 0, and 0.5. The released dataset contains 216 test records and is therefore broader than the 195-test subset emphasized in the article. The data include force amplitude, mean force, loading frequency, specimen geometry, calculated stress amplitude and mean stress, stress ratio, experimental fatigue life, failure/run-out information, measured defect-size parameters, defect depth, damage mechanism, and crack-initiation classification. The dataset can be used to validate classical mean-stress fatigue models, machine-learning fatigue-life models, and extrapolative validation strategies in which entire stress-ratio regimes are excluded from model calibration.
Files
Steps to reproduce
Thin-walled tubular specimens (ASTM E2207-02, Fig. 1) were fabricated in the vertical orientation using EOS M280 systems and MS1 powder (18Ni300 maraging steel). The recommended EOS laser powder bed fusion process parameters were applied. The specimen surfaces were cleaned of the unfused powder particles using micro-shot peening with glass beads (90–150 µm) at a flow pressure of 300 kPa. The final step was age-hardening at 490C for 6 h. The specimens were subjected to uniaxial cyclic loading at stress ratios of R = -2, -1, -0.50, -0.25, 0.0, and 0.50 using Instron testing machines (Electropuls E10000 and 8852 systems). Force-controlled constant amplitude sinusoidal loading was applied until specimen separation. The loading frequency varied between 1 and 20 Hz depending on the applied load to avoid specimen heating.
Institutions
- Opole University of TechnologyOpole Voivodeship, Opole