The role of casting defects in the fatigue behavior of notched cast aluminum alloys

Abstract

In this investigation, a number of fatigue problems related to the presence of a flow at a notch root have been addressed. The iron in front of the notch root in which the defect affects fatigue crack growth has been determined. The reductions in smooth specimen fatigue strength for constant amplitude loading, intermittent underload block loading, and for a SAE service load history caused by a flaw at notch root were determined by fatigue tests. The fatigue test data were generated on three cast aluminum materials, including Al 206, Al 319, and Al 390. The cast Al 319 material was tested in the as cast and hipped conditions. Hipping is a process in which the material is subjected to a high pressure at high temperature and then slowly cooled to eliminate or close internal flaws. Smooth specimens, notched specimens and notched specimens with an artificial or natural flaw at a notch root were fatigue tested. The smooth specimens fatigue test results showed that the block loading history, which consisted of underloads followed by constant amplitude smaller cycles, reduced the crack opening stress so that the constant amplitude cycles were fully effective. The smooth specimens results also showed that the hipping process increased the constant amplitude fatigue strength by 62.5% for the 319 cast aluminum alloy. The notched and notched with artificial or natural flaw at a notch root results, obtained under the constant amplitude loading and under the block loading history, showed that natural flaws in the as cast 319 aluminum alloy can be modeled by an equivalent drilled hole of the same size. Fatigue test results obtained under the service load history were used to study the fatigue strength reductions from the notched specimens, caused by a defect at notch root. The results showed that the fatigue lives of the 3.0 mm and 1.5 mm radius notched specimens with a 0.6 mm diameter flaw at the notch root were 40% and 38% shorter than the notched specimen fatigue lives, respectively.

The fatigue life of smooth specimens, notched specimens, and notched specimens with a flaw at the notch root subjected to constant amplitude loading, intermittent underload block loading, and a SAE service load history was predicted using a crack growth model. The crack growth model calculations were based on elastic and plastic notch strain calculations based on Neuber's formula, crack opening stress calculations, and a reference crack growth rate curve obtained during closure-free crack growth. Fatigue life predictions were in good agreement with the experimental results.

The crack growth model was used to study the effect of the variation in the flaw size and flaw position at the notch root on fatigue life. The fatigue life predictions revealed that the fatigue life of notched specimens having a small notch radius, of a size comparable to the flaw size, is more affected by a given variation in flaw size and flaw position that notched specimens having a large notch radius. The model predictions matched the observed fatigue lives. Fatigue life predictions for specimens with a flaw at a notch root using the crack growth model were compared with those obtained using conventional methods of strain-life and effective strain-life analyses. Results obtained showed that the conventional methods for fatigue life prediction using the fatigue notch factor, Kf, do not account for the decrease in concentration factor as the crack advances and cannot describe the fatigue behavior of notches with a high Kt value that decays rapidly, such as a notch with a flaw at the notch root.

In the absence of sufficient experimental data or finite element results to calibrate a model describing the variation in crack opening stress, a conservative assumption has sometimes been made that the crack opening stress throughout fatigue loading remains at the level it would have following the largest cycle. This assumption, while shown in earlier studies to be conservative, was reasonable for the materials and histories examined. It was shown in this study that this assumption is more conservative for the cast aluminum material used. The more conservativeness was caused by the crack opening stress which requires fewer cycles, after the application of an underload or an overload, to reach the steady state crack opening stress than that for the materials studied earlier (1045 Steel and 2024-T351 aluminum). The analytical crack growth model was used to determine the frequency of the cycle in the SAE GKN Grapple-Skidder history that had a crack opening stress, which when used as a constant level gives the correct fatigue life.