SUMMARY

Critical advancements in energy, transportation, and virtually every other field are enabled by the discovery of new materials with innovative and unique properties. Unfortunately, in the current development paradigm, materials are designed in a trial-and-error approach where they are iteratively generated and tested against their application requirements in an expensive and time-consuming process. Ideally, materials would be designed for their application, and different ingredients of materials science could be combined to produce the desired functionality and properties in aggregate. However, this requires a rigorous understanding of structure-property relationships across many different materials systems, which currently does not exist.
Recent materials science initiatives have highlighted the importance of using a synthesized approach of experiment and modeling to further elucidate these relationships. Numerical models should aim to robustly predict the effect of different microstructural features on material response, but certainly require validation against experimental data, ideally on several different length scales.
With these objectives in mind, using experimental in-plane deformation maps as a tool for mesoscale calibration is presented. First, an investigation of the errors associated with experimental strain maps from Digital Image Correlation (DIC) and methods for optimizing experimental and numerical protocols to reduce uncertainty are presented. Second, a method to employ in-plane strain maps in calibrating a high-order numerical model is presented highlighting the ability of the experimental dataset to further reduce the parameter space determined from experimental macroscopic load-displacement data. Lastly, a new, microstructurally-sensitive creep damage model is proposed and employed in a finite-element framework. Comparison of damage results with experimental data show excellent predictability, especially in the tertiary creep regime.