Equivalent crack size model for pre-corrosion fatigue life prediction of aluminum alloy 7075-T6. (July 2016)

• Record Type: Journal Article
• Authors: Huang, Yongfang; Ye, Xubin; Hu, Benrun; Chen, Lijie
• Is Part Of: International journal of fatigue. Volume 88(2016)
• Page Start: 217
• Page End: 226
• Publication Date: 2016-07
• Subjects: Equivalent crack size model -- Pre-corrosion fatigue -- AA7075-T6 -- Life prediction -- Correlation analysis; Materials -- Fatigue -- Periodicals; Materials -- Fatigue; Periodicals; 620.1122
• DOI: 10.1016/j.ijfatigue.2016.03.035
• Languages: English
• ISSNs: 0142-1123

Abstract

Highlights: Fatigue crack initiates from a single pit or pit cluster at the central hole. For single-crack cases ar ECS vs. ar cri is best correlated with coefficient of 0.9. For multi-crack cases ar ECS vs. r cri is best correlated with coefficient of 0.69. ECS models are developed with the best relevant metrics ar ECS ∼ ar cri and ar ECS ∼ r cri . N pre predicted with ECS models agree well with N exp with the error factor of 1.6. Abstract: Corrosion damage can significantly reduce the service life of aluminum alloy structures and endanger the structural integrity of aircraft. Here, aiming at center-hole sheet specimens of aluminum alloy 7075-T6, uniaxial fatigue tests and post-fracture analysis are performed to investigate the effect of corrosion pits on the pre-corrosion fatigue behavior. Then the best correlated parameters between corrosion pits and equivalent cracks are identified through Pearson correlation analysis. It is found that for single-crack initiations ar ECS (equivalent crack depth ∗ aspect ratio) vs. ar cri (critical pit depth ∗ aspect ratio) are best correlated with correlation coefficient of 0.9, while the best correlated parameters for multi-crack initiations are ar ECS (equivalent crack depth ∗ aspect ratio) vs. r cri (aspect ratio) with correlation coefficient of 0.69. Equivalent crack size (ECS) models are correspondingly developed with these best correlated parameters for single- and multi-crack initiations, respectively. The pre-corrosion fatigue lives predicted with our models agree well with the experimental results and the maximum error factor is about 1.6.