Date of Award


Level of Access

Campus-Only Thesis

Degree Name

Master of Science (MS)


Civil Engineering


Eric N. Landis

Second Committee Member

William Davids

Third Committee Member

Zhihe Jin


Basic crack theory has shown that the strength of brittle materials can be traced to a critical flaw in the material. With the exception of certain laboratory specimens, identifying that critical flaw is difficult in any material, but it is particularly difficult in heterogeneous materials such as concrete. In a recently developed technique, small cement-based composite specimens positioned in a load frame were scanned in-situ using X-ray microtomography. The experiments allowed internal microstructure features to be characterized before and after split cylinder fracture. In the research project described here, the 3D images resulting from the microtomographic scans were further examined to see if a critical flaw could be identified based on a combination of flaw characteristics combined with the split cylinder stress field. Two specific flaw types were examined. First, internal voids that might be starting points for cracks were examined. A stress intensity approach was developed that characterized each of the largest flaws by their position relative to the principle tensile stress and the corresponding principal direction, and the maximum area of the flaw normal to the principal direction. Through this characterization of cement paste specimens, a critical stress intensity was identified. When this critical stress intensity was applied to specimens with aggregate particles, it was only partially successful in predicting splitting load, suggesting alternate “weak links.” Thus a second flaw type was examined: the interfacial transition zone (ITZ) between the cement paste and the aggregate. This zone has long been known to affect concrete strength. The ITZ in the specimens was isolated using a number of 3D image processing routines. The focus here was on the regions of the ITZ that are aligned with porosity was proposed as the critical flaw for those specimens that failed at lower loads than the previously analyzed critical stress intensity predicted. Results showed that this critical ITZ region was not a good predictor of split cylinder, and could not by itself be considered a “weak link.” It is possible that the ITZ does act as crack initiators, but it does so in a sub critical manner such that damage must accumulate until it becomes critical.