A crack arrest methodology based on Bazant's parameter to fretting fatigue. (June 2018)

• Record Type: Journal Article
• Authors: Cardoso, R.A.; Campos, E.R.F. de S.; Ferreira, J.L.A.; Wang, D.; Araújo, J.A.
• Is Part Of: Theoretical and applied fracture mechanics. Volume 95(2018)
• Page Start: 208
• Page End: 217
• Publication Date: 2018-06
• Subjects: Fretting fatigue -- Fracture mechanics -- Short crack arrest; Fracture mechanics -- Periodicals; 620.1126
• DOI: 10.1016/j.tafmec.2018.03.001
• Languages: English
• ISSNs: 0167-8442

Abstract

Highlights: Δ K th curves influence on short crack arrest methodologies when fretting takes place. Short crack behaviour evaluation through Bazant's equation. Perpendicular cracks simplification assumption under fretting conditions. Abstract: There are several short crack methodologies available in the literature to estimate crack arrest under fretting loading conditions. Some of them claim that the most appropriate threshold stress intensity factor curve for short crack behaviour is better described by the El-Haddad curve, others assume that the approximation of the threshold curve by two straight lines in the Kitagawa-Takahashi diagram (in logarithm scale) would be a simpler and adequate threshold. Despite the non-proportional multiaxial stress field under the contact, the crack driving force in such methodologies has either been computed considering just mode I or mixed mode (I and II). The aim of this work is to conduct a review of these short crack arrest models and to propose a new and more conclusive approach. Plain fretting and fretting fatigue data available in the literature for AISI 1034, Al4%Cu and Ti-6Al-4V were considered to validate the analysis. The results showed that the accuracy of the short crack arrest methodology is strongly dependent of the threshold stress intensity factor curve considered and that the introduction of the Bazant's parameter in the analysis is preferred to obtain more sensible estimates. It was also observed that the introduction of the mode II for the computation of an effective stress intensity factor did not significantly improve the results for the data here assessed.