The objective of this work is to examine the effect of cement chemistry and physical properties on activation energy. Research efforts indicated that time dependent concrete properties such as strength, heat evolution, and thermal cracking are predictable through the concept of activation energy. Equivalent age concept, which uses the activation energy is key to such predictions. Furthermore, research has shown that Portland cement concrete properties are affected by particles size distribution, Blaine fineness, mineralogy and chemical composition. In this study, four Portland cements were used to evaluate different methods of activation energy determination based on strength and heat of hydration of paste and mortar mixtures. Moreover, equivalency of activation energy determined through strength and heat of hydration is addressed. The findings indicate that activation energy determined through strength measurements cannot be used for heat of hydration prediction. Additionally, models were proposed that are capable of predicting the activation energy for heat of hydration and strength. The proposed models incorporated the effect of cement chemistry, mineralogy, and particle size distribution in predicting activation energy.

Please read the abstract in the section 00front of this document / Thesis (PhD)--University of Pretoria, 2008. / Materials Science and Metallurgical Engineering / PhD / Unrestricted

Being able to extrapolate interaction values from a database of pullout resistance testing results may possibly help with narrowing down the most suitable reinforcement/fill material combinations for a Mechanically Stabilized Earth wall, thereby reducing the number of tests needed for a design and maximizing the efficiency of the system. The objectives of this thesis include the following: collect and organize a broad collection of data in a way that can assist in preliminary selection of interaction properties for uniaxial geogrids; analyze the collection of data for trends related to geogrid polymer type; analyze the collection of data for trends related to the presence of fines in the fill material; compare the collected data to previous studies on the effects of geogrid specimen length on pullout performance; and compare the collected data to previous studies on the effect of geogrid rib thickness to mean particle size ratio on normalized bearing stress and CI values. The data from 101 pullout tests are presented in tabular and graphic form so that the coefficient of interaction may be interpolated for many geogrid/fill material combinations. The effect of polymer type (PET vs HDPE) was shown to have little effect on how a geogrid performs in a fill material. In one case, the two polymer types exhibit differing trends within the same fill material. The presence of fines (>12% by weight) in the fill material results in a significant decrease in the coefficient of interaction when compared to clean granular fills. The effects of geogrid embedment length have significant effects on the results of geogrid pullout tests. Samples with shorter lengths were shown to carry a greater load per unit area than longer samples. Normalized bearing stress is shown to be heavily influenced by the geogrid transverse rib thickness to mean particle size ratio (B/D50). For a particular fill material, normalized bearing stress decreases linearly with increasing B/D50. For a particular geogrid, normalized bearing stress is shown to have a bi-linear behavior with increasing B/D50. Initially, normalized bearing stress increases with increasing B/D50. After reaching a peak, normalized bearing stress begins to decrease with increasing B/D50. / text