Competing fatigue mechanisms in Ni-Base superalloy Rene 88DT

Nickel base superalloys exhibit superior high temperature mechanical properties required for aircraft engine components. It has been known that the processing of these alloys by the powder metallurgy route introduces inclusions inside the material. The presence of such inclusions often leads to competing failure modes in fatigue that is described by a step-wise or two distinct S-N curves involving both the surface and internally-initiated cracks, resulting in large uncertainties of fatigue life. A clear understanding of such behavior is yet to be established. The principal objective of this research is to examine the effect of inclusions on the extent of fatigue failure competition from surface and internal initiators at two different specimen test volumes. Experimental fatigue testing has been performed to explore how the presence of inclusions affects the competing fatigue failure modes. In addition, how the competing failure modes will behave with changes in the specimen size was also studied. Two groups of material each with two different specimen sizes were used in this study. It has been shown that the two crack initiation mechanisms occurred in the small unseeded René 88DT specimens tested at 650ºC over the stress range tested. Additionally, the fatigue lives were reduced with increase in specimen volume. All fatigue failures in seeded material occurred due to crack initiations from the seeded inclusions. In the fatigue life of seeded material, two competing and separate S-N curves were found in small test volume, whereas, in the large test volume, the regions were separated by a "step" in S-N curve. It has been found that the largest inclusion size observed in metallographic surfaces was smaller than the size determined from the fatigue failure origin. An analysis method based on extreme value statistics developed by Murakami was used to predict the largest size of inclusion in the test volume. The results of this study clearly show that competition for fatigue failure between surface and internal initiations is affected by inclusion content and specimen volume. Extreme value statistics is capable of predicting the largest inclusion that can start fatigue failure, even if the inclusion is not seen in microscopy.