Comparison of a Datum Temperature Calibration Method With Traditional Approach for Norton Power Law

Abstract

Abstract Methods for minimum-creep-strain-rate prediction have evolved. Many models have been proposed, and different calibration techniques are used. Often the limitation of these models for accurate prediction arises due to a lack of long-range data incorporating both the low-stress and high-stress regions. This problem is more prominent for novel materials with very little data and may require long-term creep tests, delaying the material’s qualification. Model calibration against short-range data may lead to an inflection during extrapolation. In this study, a datum temperature (DT) calibration method derived from Parametric Numerical Isothermal Datum (P-NID) is compared with the traditional calibration approach for minimum-creep-strain-rate prediction using Norton power law. Minimum-creep-strain-rate data for Inconel 617 at five temperature levels (800 to 1000°C) and stress ranging from 11 to 122 MPa are used. Two different forms of the Norton power law are calibrated using the traditional approach and the most suitable form for Inconel 617 is selected. Next, the model is calibrated using the datum temperature calibration approach. In the datum temperature method, the data from different temperatures are transferred to a datum temperature creating a wide range of parametric data followed by model calibration against the transferred data at datum temperature. Finally, the model is transferred back to the original temperatures. The traditional approach and datum temperature method results are compared in terms of accuracy, calibration techniques, extrapolation, and limitations for Inconel 617. The datum temperature method is found to be accurate, like the traditional approach, however, requires comparatively less effort during calibration since the model is calibrated against a single temperature instead of multiple temperatures. Thus, the material constants are independent of temperature and stress resulting in stable, inflection-free, and reliable extrapolation over the traditional approach. A step-by-step procedure is provided to derive the datum temperature transformation equations and the calibration method. Finally, a general guideline is provided to apply the datum temperature method to any existing models.