Study of the effect of defects on fatigue life prediction of additive manufactured Ti-6Al-4V by combined use of micro-computed tomography and fracture-mechanics-based simulation

The significant variability in the high-cycle fatigue of additive manufactured (AM) Ti-6Al-4V has impeded its confident application in fatigue-critical components. One of the major factors responsible for the fatigue-life variability is the presence of intrinsic defects left by the additive manufacturing process. The goal of this work is to investigate the fatigue-life variability as the function of (i) the locations of the defects, viz., internally or on the surface, (ii) the distance to the free surface if it is an internal defect, (iii) the sizes of defects and (iv) the accurate stress at the defect locations. In this study, AM Ti-6Al-4V specimens were firstly scanned by micro-computed tomography, then tested in air under high-cycle fatigue with stereo-digital image correlation, and finally examined post-mortem using scanning electron microscopy. Defects detected by micro-computed tomography have been mapped with the local stress reproduced from the stereo-digital image correlation measurement and correlated with the defects found in post-mortem fractography by scanning electron microscope. Wide variation in the fatigue life was indeed found from the tests and more than half of the tested specimens initiated cracks from the internal defects. A fracture mechanics-based approach was developed to account not only for the defect sizes but also for different crack growth rates between surface cracks in air and internal cracks in vacuum. In comparison with the nominal stress-based approach, the scatter of the fatigue life prediction can be significantly reduced from ± 60 times to ± 6 times by applying this fracture mechanics-based methodology.