Mechanistic Model to Predict Statistical Scatter in Fatigue Response of Titanium Aluminide

Abstract

This paper discusses research to develop a computer modeling technique to predict the statistical behavior of fatigue in fully lamellar TiAl. Probabilistic micromechanics was used to relate the random nature of the microstructure to the fatigue response of smooth round bar specimens. The effort showed that models proposed in the literature addressing the microstructural mechanisms that govern fatigue damage accumulation could be combined with the concept of mesomechanics to properly account for the characteristic sizes and multiple phases of fatigue. The capability developed can be used to virtually test large numbers of specimens for which actual laboratory testing would be cost prohibitive.