Probabilistic analysis of aircraft gas turbine disk life and reliability

Two series of low-cycle-fatigue (LCF) test data for two different groups of aircraft gas turbine engine compressor disk geometries were reanalyzed and compared by using Weibull statistics. Both groups of disks were manufactured from titanium (Ti-6AI-4V) alloy. A probabilistic computer code called Probable Cause was used to predict disk life and reliability. A material-life factor A was determined for the titanium (Ti-6AI-4V) alloy by using fatigue disk data and was applied to predict disk life as a function of speed. A comparison was made with results from the currently used life prediction method, which is based on crack growth rate. Applying an endurance limit to the computer code did not significantly affect the predicted lives under engine operating conditions. Failure location predictions correlated with experimentally observed failure locations in the LCF tests. A reasonable correlation was obtained between the disk lives predicted by using the Probable Cause code and those predicted by using a modified crack growth method. Both methods slightly overpredicted life for one disk group and significantly underpredicted it for the other.