An adaptive phase field model for high cycle fatigue analysis across specimen geometries

High cycle fatigue behavior of components with notches and cracks, which are common in engineering applications, is significantly influenced by geometrical size effect and complicates accurate fatigue life prediction. Traditional methods for addressing this issue often limited to specific types of components and rely heavily on empirical correlations. To overcome these limitations, an adaptive phase-field model was proposed in this work, which integrates a stress triaxiality-dependent fatigue degradation function, a cycle-time transfer scheme, and a thermal analogy solution strategy. This model allows consistent quantification of geometrical size effect across various specimen geometries and notch configurations. The model's effectiveness was validated using a wide range of experimental datasets, demonstrating its ability to accurately capture the fatigue behavior of both notched and cracked specimens. Notably, the model achieves good computational efficiency, completing one million cycles within 16 h, and requires minimal calibration with only a few experimental data pairs. Additionally, the fatigue damage in the model is designed to depend solely on stress triaxiality and strain energy at a material point, making its application independent of the presence of notches or cracks. These features suggest that the proposed model can serve as a promising numerical tool for fatigue life management and structural design of components.