| Abstract Scope |
Oxide inclusions and porosity are present in nickel alloys that are made using powder metallurgy. Both features can nucleate fatigue cracks. Surface oxide inclusions can significantly reduce fatigue lives. This is evident if lives are compared to those for crystallographic cracks that develop from slip bands, which are characterized by facets on fracture surfaces. Pores mainly occur from trapped gas in powder particles, whereas inclusions typically arise from melting the alloy and from powder manufacture. Since most inclusions are too small to be detected from ultrasonic inspection, probabilistic lifing assessment is required to ensure the risk of failure from “melt anomalies” in disk rotors is acceptably low. Such calculations apply a mathematical description of inclusion size and frequency. This paper examines methods for characterizing size and frequency of oxide inclusions and porosity from powder and billet material, which includes material that contains oxide inclusions or “seeds” that were added intentionally to understand fatigue behavior. Large bar tensile (LBT) testing of billet material was found to be the most capable method for characterizing inclusion size. However, rate of occurrence information cannot be measured directly; it must be implied by fitting inclusion size data to probability functions. Given uncertainties with this approach, a method has been devised to measure rate of occurrence directly. Inspections of polished surfaces of billet material have been shown to be viable for production use. Porosity of a coarse powder fraction has been characterized by mounting powder in resin so that automated image capture and analyses could be undertaken on polished surfaces. Finally, the sizes of the very low frequency of larger inclusions, which are detected by ultrasonic inspection of billet, have been characterized from electron microscopy on polished surfaces. |