Rapid estimation of spectrum crack-growth life based on the palmgren-miner rule

Abstract

Failure investigations of structural cracking require a crack-growth life prediction method which can be rapidly applied and which directly displays the sensitivity of life to the crack configuration and the random cyclic stress environment. Such predictions are based on first-order nonlinear rate equations which are fitted to material properties measured under constant-amplitude stress. Many digital programs have been developed to simulate cycle-by-cycle crack growth based on these rate equations. This paper outlines a method better suited to rapid assessment. The method is based on the Palmgren-Miner Rule (linear damage summation), which was originally proposed for predicting crack-initiation life from empirical data. The Palmgren-Miner Rule is shown to be exact for crack growth when the stress and crack-length variables in the rate equation are separable. Such applications are limited, but a numerical experiment shows that the Palmgren-Miner Rule also produces acceptably accurate estimates in practical cases when the rate equation embodies commonly observed stress sequence effects. In its present state of development, however, the method cannot be applied to interaction phenomena, i.e. situations where the stress sequence can change the rate equation.