| Fatigue crack propagation tests were performed on the polycrystalline, powder metallurgy Ni-base superalloy, KM4. Two different heat treatments, producing grain sizes of 6 μm and 55 μm, were investigated. Tests were conducted at load ratios ranging from R = 0.3 to R = 0.7 at two different frequencies, 100 and 1000 Hz. Fatigue crack propagation behavior was studied at 20°C, 550°C and 650°C. Intermediate growth rate observations showed results similar to those observed by other researchers at frequencies lower than those employed in this study. The general observations showed that increasing the grain size, decreasing the load ratio, decreasing the temperature and increasing the frequency all had the effect of increasing the fatigue crack propagation resistance at intermediate growth rates. Threshold FCP behavior showed a much more complicated dependence on load ratio, grain size, frequency and temperature. In some cases, increased frequency resulted in decreased FCP threshold while in other cases, it produced the opposite effect. This complex behavior can, in part, be attributed to a transition in the mode of failure from transgranular to intergranular, however, analysis also revealed that system variable interactions (for instance frequency/temperature interactions or grain size/temperature interactions) must also be accounted for in order to understand the complex threshold behavior.; Optical profilometry was utilized to obtain a quantitative assessment of the fracture surface roughness in an attempt to correlate measured threshold values. Typical roughness parameters such as RL, R a, and Rq (or RMS) were studied. An alternate parameter, the average microscopic slope ( m), was also investigated. No direct correlation was observed between fracture surface roughness, as defined by these parameters, and measured threshold. Compliance measurements did, however, reveal the presence of fatigue crack closure for some conditions at room temperature and 550°C. It was speculated that the observance of closure, mainly at room temperature, was the result of roughness-induced closure processes acting in the crack wake. Further assessment of the roughness parameters revealed that average microscopic slope was useful for assessing potential reasons for roughness crack closure. This parameter was used in conjunction with crack wake opening displacement information obtained via finite element analysis to predict the amount of Mode II displacement required to induce closure. The required Mode II offsets predicted by the model were consistent with scanning electron microscopy observations obtained on interrupted tests.; Characterization of oxide films at 550°C resulted in scale thicknesses which by themselves could not explain the elevated temperature closure response based on oxide wedging mechanisms. A simple model was employed to adjust the crack wake opening displacements for the growing oxide films in an attempt to explain the elevated temperature behavior as a combined effect, the results of which were inconclusive. Thus, closure models could only partly explain the threshold FCP behavior indicating that intrinsic mechanisms dominate the behavior of KM4 superalloys. It was concluded that the complexity of the threshold behavior in this material was a result of a combination of both intrinsic and extrinsic mechanisms. |
