Mechanical & Materials Engineering, Department of


Date of this Version



Published in Journal of Materials Science 41 (2006), pp. 6510–6519; doi 10.1007/s10853-006-0185-6 Copyright © 2006 Springer Science+Business Media, LLC. Used by permission.


A model is presented herein for predicting the evolution of numerous cracks on multiple length scales, the objective of such a model being to develop the capability to predict failure of structural components to perform their intended tasks. Such a capability would then be useful as a predictive tool for designing structural components so as not to fail, but rather to succeed in performing their intended tasks. The model developed herein is somewhat involved, being based in continuum mechanics and thermodynamics, but is nevertheless expected to be cost effective (wherever sufficient accuracy permits) when compared to more costly experimental means of determining component life. An essential ingredient within the context of the model is that cracks must develop on widely differing length scales. Where this is observed to occur in nature, which is surprisingly often, there are potential simplifications over more generally described but practically untenable approaches, that can lead to (at least partly) computational multiscale algorithms capable of assimilating failure due to multiple cracking with a high degree of accuracy. The model presented herein will be briefly described within a mathematical framework, and an example problem will be presented that is representative of certain currently relevant technologies.