No abstract

A study was conducted to determine the processes which govern hold time crack growth behavior in the LSHR disk P/M superalloy. Nineteen different heat treatments of this alloy were evaluated by systematically controlling the cooling rate from the supersolvus solutioning step and applying various single and double step aging treatments. The resulting hold time crack growth rates varied by more than two orders of magnitude. It was shown that the associated stress relaxation behavior for these heat treatments was closely correlated with the crack growth behavior. As stress relaxation increased, the hold time crack growth resistance was also increased.The size of the tertiary γ' in the general microstructure was found to be the key microstructural variable controlling both the hold time crack growth behavior and stress relaxation. No relationship between the presence of grain boundary M 23 C 6 carbides and hold time crack growth was identified which further brings into question the importance of the grain boundary phases in determining hold time crack growth behavior.

Isothermal, in-phase and out-of-phase axial-torsional fatigue experiments have been conducted at 760°C on uniform gage section, thin-walled tubular specimens of a wrought cobalt-base superalloy, Haynes 188. Test control and data acquisition were accomplished with a minicomputer. Fatigue lives of the in- and out-of-phase axial-torsional fatigue tests have been estimated with four different multiaxial fatigue life prediction models, a majority of which were developed primarily for predicting axial-torsional fatigue lives at room temperature. The models investigated were: (1) the von Mises equivalent strain range, (2) the modified multiaxiality factor approach, (3) the modified Smith-Watson-Topper parameter, and (4) the critical shear plane method of Fatemi, Socie, and Kurath. In general, life predictions by the von Mises equivalent strain range model were within a factor of 2 for a majority of the tests, and the predictions by the modified multiaxiality factor approach were within a factor of 2, while predictions of the modified Smith-Watson-Topper parameter and of the critical shear plane method of Fatemi, Socie, and Kurath were unconservative and conservative, respectively, by up to factors of 4. In some of the specimens tested under combined axial-torsional loading conditions, fatigue cracks initiated near extensometer indentations. Two design modifications have been proposed to the thin-walled tubular specimen to overcome this problem.

The 1100 °C cyclic oxidation performance of 25 Ni-base commercial and developmental alloys was compiled from an extensive database and ranked according to the 200 h weight change. Cyclic oxidation performance of superalloys is directly controlled by composition. These conventionally cast superalloys were composed of base elements [Ni-Co-Cr-Al], refractory elements [Nb-Mo-Ta-W], oxygen-active elements [Ti-Zr-Hf], light elements [B,C], and occasionally [V-Mn-Si], with P and S trace impurities. The oxidation results were broadly categorised as less than 4 mg/cm 2 weight loss for alloys with high 5-6% Al and 3-9% Ta, and with low ≤ 1% Ti (wt.%). Conversely, weight loss of 200-300 mg/ cm 2 characterised alloys containing low < 3.5% Al, no Ta, and high > 3% Ti. These trends correlated with beneficial and detrimental scale phases previously reported. An unambiguous Cr effect was masked because of its strongly coupled, but inverse, correlation with Al. Multiple linear regression was used to fit alloy composition to a simple logarithmic weight change transform. The function contained 10 terms and yielded a correlation coefficient, r 2 , of 0.84. Various graphical representations helped to further illustrate, quantify, and predict complex oxidation effects within a 10-element compositional space.

No abstract

No abstract

No abstract

The results are reported for high-temperature axial and torsional low-cycle fatigue experiments performed at 760° C in air on thin-walled tubular specimens of Haynes 188, a wrought cobalt-base superalloy. Data are also presented for mean coefficient of thermal expansion, elastic modulus, and shear modulus at various temperatures from room to 1000° C, and monotonic and cyclic stress-strain curves in tension and in shear at 760° C. This data set is used to evaluate several multiaxial fatigue life models (most were originally developed for room temperature multiaxial life prediction) including von Mises equivalent strain range (ASME Boiler and Pressure Vessel Code), Manson-Halford, Modified Multiaxiality Factor (proposed in this paper). Modified Smith-Watson-Topper, and Fatemi-Socie-Kurath. At von Mises equivalent strain ranges (the torsional strain range divided by 3, taking the Poisson’s ratio to be 0.5), torsionally strained specimens lasted, on average, factors of 2 to 3 times longer than axially strained specimens. The Modified Multiaxiality Factor approach shows promise as a useful method of estimating torsional fatigue life from axial fatigue data at high temperatures. Several difficulties arose with the specimen geometry and extensometry used in these experiments. Cracking at extensometer probe indentations was a problem at smaller strain ranges. Also, as the largest axial and torsional strain range fatigue tests neared completion, a small amount of specimen buckling was observed.