Multiregion analysis of creep rupture data of 316 stainless steel

A creep rupture data set of 316 stainless steel containing 319 data points at nine heats was subjected to a conventional single-region analysis and a multiregion analysis. In the former, the conventional Larson–Miller analysis was applied to the whole data set. In the latter, a data set of a single heat is divided into several data sets, so that the Orr–Sherby–Dorn (OSD) constant Q takes a unique value in each data set, and the conventional OSD analysis was applied to each divided data set. A region with a low value of Q appears in long-term creep of eight heats. Predicted values of the 10 5 h creep rupture stress of three heats were lower than the 99% confidence limit evaluated by the single-region analysis, suggesting that the single-region analysis is error prone. The multiregion analysis is necessary for the correct evaluation of the long-term creep properties of 316 stainless steel.