- Published 2013

Accurate determination of constitutive modeling constants used in high value components, especially in electric power generation equipment, is vital for related design activities. Parts under creep are replaced after extensive deformation is reached, so models, such as the Norton-Bailey power law, support service life prediction and repair/replacement decisions. For high fidelity calculations, experimentally acquired creep data must be accurately regressed over a variety of temperature, stress, and time combinations. If these constants are not precise, then engineers could be potentially replacing components with lives that have been fractionally exhausted, or conversely, allowing components to operate that have already been exhausted. By manipulating the Norton-Bailey law and utilizing bivariate power-law statistical regression, a novel method is introduced to precisely calculate creep constants over a variety of sets of data. The limits of the approach are explored numerically and analytically. INTRODUCTION Material selection is a critical stage in mechanical design engineering of structural components. Perhaps the most important consideration for parts subjected to long term use are expected service life, acceptable deformation rate, and the environment in which the material will be used. In order to accurately determine creep rupture life, engineers use analytical approaches to simulate the primary and secondary creep response. An example of such a model is the Norton-Bailey model, which contains three temperature dependent regression constants. The methods used to *corresponding author: apg@ucf.edu D. S. Segletes Siemens Energy 5101 Westinghouse Blvd, Charlotte NC 28273 optimize these constants typically involve manual curve-fitting to creep data in order to acquire best fits across several creep curves. If the constants found were their true values, then plotting the Norton-Bailey values versus time would result in a near-perfect match of the data. In some situations, the constant determination is hampered by sparse data sets at intervals of strain (e.g. 0.1%, 0.5%) or at constants times (1 hr, 10 hr). Research was conducted to develop a formulation to identify power law creep constants that would result in an optimal fit with creep data across test variables of both stress and temperature. The purpose of this investigation is to develop a reliable approach to regressing multivariate power law type data. A background look at creep deformation, other creep models, and general approaches to constant determination are discussed next. Following that, the methods being investigated are derived and tested on both physical and simulated data and its limitations are discussed. CREEP DEFORMATION Constitutive models have been developed to interpolate and predict the deformation behavior of materials exhibiting time-dependent, inelastic deformation. A model commonly applied for the primary and secondary creep regimes was developed by Bailey and Norton [1], i.e.,

@inproceedings{May2013TheAO,
title={The Application of the Norton-bailey Law for Creep Prediction through Power Law Regression},
author={Dustin May and Ali P. Gordon and D. S. Segletes},
year={2013}
}