A rapid method for determining service fatigue strength data is to test the product or device at an elevated load. Load-life relationships for engineering materials normally show a large reduction in life for a small increase in load. Unfortunately, there is little agreement on the exact nature of this load-life relationship. Some sources show a linear semi-log relationship. Some show a single linear log-log relationship. And some show a two-line log-log relationship with the shallower slope line active after ten-million load cycles or more. The two-line relationship agrees more closely with expected behavior. It predicts a difference between the load-life characteristics in the field at moderate service loads and the load-life characteristics in the laboratory with high test loads. However, the determination of the location and slope of the second, shallow, high-life line is an expensive proposition. In this paper, a parabolic log-log load-life model which predicts a gradual transition from the short-life testing environment to the long-life use environment is presented and compared to the two linear models. Fitting the model to test data is described as is the use of the model in planning tests for failure prevention prediction.

This content is only available via PDF.
You do not currently have access to this content.