Graduate Assistant, University of Cincinnati, Cincinnati, Ohio
Pages: 30 Published: Jan 1982
Four models for the creep-fatigue interaction were evaluated for their ability to predict fatigue behavior at 650°C (1200°F) of thermomechanically processed René 95, a nickel-base superalloy used for turbine disks. These models are the Strain-range-Partitioning Model, the Frequency-Separation Model, the Frequency-Modified Damage Function, and the Damage-Rate Model. The data base was divided into baseline and verification tests. The correlations of the baseline tests, having the smallest standard deviation of 0.18, were by the Frequency-Separation Model and the Frequency-Modified Damage Function. The other two models had standard deviations greater than 0.22. These standard deviations correspond to worst correlations of 3.6 and over 5.0 times the observed life. The ratios of the observed to the predicted life of the verification tests ranged from 5.2 by the Frequency-Modified Damage Function to 18 by the Frequency-Separation Model. Each of the models consistently overpredicted or underpredicted the lives of certain types of tests in the data base. The models did not account for the mean stresses observed in the fatigue tests. A method based upon the stress versus minimum creep rate and the cyclic stress-strain curves is suggested to indicate whether an environmental-fatigue interaction or a creep-environmental-fatigue interaction is present. For René 95, this method indicates an environmental-fatigue interaction.
René 95, crack initiation, low-cycle fatigue, high-temperature fatigue, strain-rate effects, nickel-base superalloys, creep-fatigue interaction
Paper ID: STP32425S