Statistical evaluation of the failure-relevance of inhomogeneities and defects on crack initiation in the VHCF regimeTuesday (27.09.2016) 12:00 - 12:15 Part of:
In the present work the influence of microstructural discontinuities on fatigue lives and hence their failure relevance in the VHCF-range was investigated. For this purpose fatigue tests accompanied by comprehensive microstructural analyses were carried out in order to determine typical crack initiation sites and related damage mechanisms. The investigations focused on two reference materials, of which one can be classified as quasi defect-free while for the second material the existence of defects such as non-metallic inclusions has to be considered. The nickel-based superalloy Nimonic 80A (quasi defect-free) and the metastable austenitic stainless steel 1.4301 with a high volume fraction of deformation-induced ?’ martensite (defect-afflicted) were investigated. The effects of typical crack-initiating microstructural discontinuities were modeled by corresponding failure-relevant parameters. The stress concentrations at crack initiating twin boundaries and at regular grain boundaries in Nimonic 80A were quantified using a misorientation factor by Blochwitz et al. (1997) resulting in a crack initiation parameter. For the defect-afflicted 1.4301 the effect of size and location of non-metallic inclusions was evaluated by means of a stress intensity factor reflecting the local stress at such types of defects. By analyzing the distribution of failure-relevant parameters in fatigued specimens it could be shown that crack initiation predominately occurs at defects with the maximum values of the respective parameter. Hence, a statistical model for the correlation between the failure-relevant parameters and corresponding number of cycles until failure or crack initiation was developed. The analysis and statistical modeling of the defined damage-relevant defects was carried out on the basis of comprehensive metallographic investigations of the reference materials in the as-received condition. Extreme value statistics were used to model the size and (if necessary) space distributions of the larger values of defined damage-relevant discontinuities/defects. The model was applied in order to predict fatigue life of specimens tested within the present work and in addition, was compared with fatigue test results from other researchers.