Collaborative Research: An Integrated Microstructure-Based Approach to Property Prediction for Cement-Based Materials

The project is a collaborative effort between the research groups at the University of Maine and the University of California, Davis. It represents a convergence of 3D micro-structural imaging with discrete element computational modeling to address the need for quantitative links between salient micro-structural properties and bulk material performance. High-resolution 3D images of material microstructure will be produced using x-ray microtomography (XMT). These images will be used to examine the micromechanical response of cement composites to mechanical loading and drying shrinkage. 3D deformation fields and internal crack distributions will be measured at selected loading stages and correlated with the spatial distribution of inclusions and local variations in cement paste porosity. Through developments of a computational model and appropriate inverse analyses, we will provide refined estimates of micromechanical properties that have been difficult to measure, including cement-aggregate interfacial strength, bridging forces, and eigen stresses produced by drying shrinkage. The computational models will be based on an explicit representation of micro-structural features, as determined from the 3D image data. Whereas most of the measurements and analyses will be conducted at the micron to centimeter scale, the results will be interpreted as one of the several critical length scales necessary for modeling practical systems. The project results will contribute to the understanding of concrete fracture at the scale in which fine aggregates and meso-pores can be viewed as discrete entities. Little quantitative work exists for this situation. Knowledge of the details of cracking at this scale is essential to improving the performance of concrete materials, including current efforts to extend the service life of concrete structures exposed to severe environments. The proposed work fits within a paradigm shift towards morphological bases for material modeling. As such, the proposed work is relevant to a wider range of material classes, beyond cement and concrete composites