Comparison of two ‘heat correction’ methods for simultaneously correlating multiple heat stress rupture data

The heat corrections (HC) used in two multiple heat stress rupture correlations, one from Manson and Park (MP) and the other from the present authors (RSR), are compared as to their efficacies in correlating multiple-heat stress rupture data for a 0.5Mo–0.15C steel, a 9Cr–1Mo steel, two grades of 2.25Cr–1Mo steels, and two grades of AISI type 316 stainless steels. Two types of fits of isothermal multi-heat data were carried out; in one case the curvatures of the isothermal plots of log(rupture life) against log(stress) were restricted to be non-positive (to be consistent with the behaviour expected from model-based considerations), while in the other case no restrictions were imposed on the possible curvature variations, as is common in the creep literature. Irrespective of whether the curvature restrictions were imposed or not, the HC adopted in RSR proved superior to that in MP for all these materials, though the extent of superiority is apparently dependent on the material. Interestingly, in most cases, the imposition of the curvature restrictions only marginally affected the quality of fit, though this may enhance the regime of the correlation that can be used for extrapolation.