For Grade 92 steel (9Cr-0.5Mo-1.8W-V-Nb), multi-batch stress-rupture measurements are shown to be rationalized through relationships which involve only the activation energy for matrix diffusion (300 kJmol−1) and the ultimate tensile stress values at the creep temperatures. The resulting ‘master curve’ is at least as impressive as those obtained using traditional parametric methods, but with the empirical parameters replaced by physically-meaningful properties. These approaches lead to straightforward procedures for extended extrapolation of short-term data, with analyses of test results for creep lives less than 5000 hours predicting 100,000-hour rupture strengths. Indeed, noting that the allowable creep strengths for Gr. 92 steel have been reduced progressively as longer-term fracture data have become available, the present predictions coincide well with the lower limits of the most recent estimates determined for stress-temperature combinations producing failure in times up to 100,000 hours and more. Validation of the new methodologies can be achieved by independent analyses of standard property sets for other creep-resistant steels, introducing the prospect of a marked reduction in the scale and costs of the experimental programs currently undertaken to provide long-term engineering design data.
