Modification of the Murakami fatigue crack growth model for multiaxial loading: Application to tensile-dominated failure of small defects

While small defects are ubiquitous in metals and act as critical sites for fatigue failure, high-fidelity life assessment using fracture mechanics remains a challenge due to the multiscale nature and complex microstructural mechanisms involved in the damage process. Despite these difficulties, this work develops an easy-to-use yet effective life prediction model for defect-containing metals under multiaxial loading. The proposed formulation extends the small-crack-based Murakami model, originally restricted to uniaxial stress conditions, by incorporating a Walker-type fatigue parameter and a critical plane approach to account for mean stress and multiaxial stress-strain effects. New high-cycle fatigue data for 1045 carbon steel specimens with artificial cylindrical surface defects (400 μm diameter) were obtained under various proportional and nonproportional axial-torsional loadings. The investigation focused on the mean stress effect on small fatigue crack growth from these defects. Fatigue lives ranged from 10 5 to 10 7 cycles (run-out condition). The model constants were calibrated using uniaxial test data, and the model was evaluated against the multiaxial test results. The proposed model provided a reasonable correlation with experimental fatigue lives, with most predictions falling within factor-of-four scatter bands. Moreover, it accurately predicted the observed small-crack angles, indicating that the failure mode under the investigated conditions is dominated by Mode I (tensile) small-crack growth.