Published Online: 1 February 2004
Page Count: 12
Graduate Research Assistant, School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, GA
Professor, School of Materials Science and George P. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA
(Received 8 November 2002; accepted 11 August 2003)
Directional solidification of nickel-base superalloy components for gas turbine rotor blades offers a significant improvement in creep strength while at the same time yielding an anisotropic, heterogeneous microstructure. In such a material, some of the continuum assumptions of linear elastic fracture mechanics begin to break down, and local (microscopic) conditions can lead to significant deviation in fatigue crack growth rate from the global (macroscopic) trend. Fatigue crack life prediction for a large population of fielded components requires a probabilistic treatment of the material fatigue crack growth behavior. A common approach to probabilistic fatigue crack life prediction involves sampling the Paris law coefficients from a large number of crack growth experiments, which can lead to effectively “smoothing” the local intraspecimen variability out of the model. The length scale of variability is discussed as it relates to material microstructure and crack life prediction. Results from fatigue crack growth experiments on a directionally solidified superalloy are presented and spatial variation in the fatigue crack growth rate is examined. Periodicity of the crack growth rate variation is compared with the scale of microstructural heterogeneity.
Paper ID: JAI11565