Fatigue Life Estimates for a Notched Member in a Corrosive Environment

Abstract

It is often assumed that the effects of an aggressive environment can be included in fatigue life estimation procedures by determining the material properties in the environment and at the frequency of interest. An analytical and experimental program was conducted to confirm or refute this assumption. Automotive grade aluminum alloy, 5454-H32, in 3 percent NaCl solution and laboratory environment was selected for this study. A simple model where the total fatigue life is the summation of the portion where fatigue damage is best described by the notch strain field, and the portion where nominal stress and crack length dominate damage assessment, was used to estimate fatigue lives for center notched plates. Smooth cylindrical specimens were employed to determine the material properties for initiation. The environment had a large influence on the initiation resistance of this material at long fatigue lives, whereas at shorter fatigue lives (i.e., <104 cycles) there was little effect. Center cracked plates were used to determine the crack growth rates. Linear elastic fracture mechanics concepts were employed to estimate crack propagation lives. Approximately a factor of three reduction in crack propagation life was attributable to the hostile environment. Center notched plate specimens with K t = 2.4 and K t =5.1 were tested in both environments to examine the model. The accuracy of the fatigue life predictions in relation to the experimental data were comparable in 3 percent NaCl solution to the results obtained in laboratory air.