Improvements in Empirical Representation of A356-T6 Cast Aluminum Alloy Round-Robin Low Cycle Fatigue Data

The results of the SAEFDE Committee's round robin low cycle fatigue test program with A356-T6 cast aluminum alloy indicated that the conventional low cycle fatigue model was not a satisfactory representation of the data. This occurred because the elastic strain amplitude-life curve was not log-log linear and this yielded a non-conservative fatigue life representation at both extremes of long and short lives. This paper involves a reanalysis of the A356-T6 composite all-laboratory data using two additional empirical models. These models are:

1. 1.
   linear log-log total strain amplitude-life
2. 2.
   bilinear log-log elastic strain amplitude-life

Both proposed empirical models improve the representation of the data compared to the conventional low cycle fatigue model. The bi-linear log-log elastic equation, however, when added to the plastic equation, yields a discontinuous curve with non-conservatism in the region of the discontinuity. Of the models examined, the linear log-log total strain amplitude model provides the best representation of the data. Both proposed models could easily be incorporated into a local notch strain life prediction program, however no experimental verification has been performed to indicate that the local notch strain variable amplitude life prediction method is applicable to cast aluminum alloys.