A Dynamic Optical Correlation Technique For Fatigue Failure Prediction

Abstract

A continuous optical correlation technique for fatigue damage detection and fatigue failure prediction in metals is described. Information about the metal surface is recorded on a thermoplastic-photoconductor device and compared with the actual surface as the fatigue test proceeds. Results are presented as plots of optical correlation intensity (CI) versus number of fatigue cycles (N) for aluminum 2024-T3 sheet. A decelerating loss of log CI versus N is observed over about the first 30 percent of fatigue life, a linear loss from 30 to 80 percent approximately, and an accelerating loss over the final 20 percent of life. The correlation intensity curve provides a sensitive indication of fatigue damage occurring in the specimen, and can be used to monitor crack formation and to predict impending failure.