Exploring Microcracking in Slag-Based Alkali-Activated Cements

Mihalik, Jarred [Browse]
Senior thesis
76 pages


White, Claire [Browse]
Princeton University. Department of Civil and Environmental Engineering [Browse]
Class year
Summary note
Concrete's importance as a construction material cannot be overstated, but it is far from a perfect material. The production of one ton of ordinary Portland cement (OPC) releases nearly a ton of carbon dioxide into the atmosphere. Alkali-activated cements (AAC) can lower this figure by reducing the amount of OPC used in concrete or replacing it entirely. One type of AAC uses a combination of ground granulated blast-furnace slag (GGBS) and an alkaline solution to replace the OPC found in standard concrete mixes, but it is also prone to microcracking that occurs within 24 hours of the AAC specimens being demolded. Here, an investigations was carried out to determine possible causes of microcracking by varying the AAC composition and observing the subsequent changes. The factors assessed in the experiment included different chemical activators (NaOH, Na2SiO3, and Na2CO3), different sources of GGBS (Zeobond, AUS and Holcim, USA), and different atmospheric drying conditions (ambient laboratory conditions and N2). For each AAC sample, the crack area, mass loss during drying, and pore characteristics were measured. It was found that cements made with a sodium hydroxide activator experienced the least amount of cracking, followed by sodium carbonate and sodium metasilicate. The impact of variations in atmospheric conditions and the type of GGBS trends were much smaller, with Holcim cement cracking slightly more than Zeobond cement and nitrogen dried cement cracking slightly more than cement dried under a normal atmosphere. However, the mixes using sodium carbonate experienced the most mass loss, followed by sodium metasilicate and sodium hydroxide respectively. These results indicate that the chemical activator has the greatest effect on a specimen's tendency to crack, and also reveal that mass loss is not directly correlated with crack amount. Instead, it is seen that there is a direct correlation between surface cracking and the average pore size measured using mercury intrusion porosimetry, as sodium metasilicate samples had the smallest pore size and smallest overall porosity. These results are representative of the current understanding of how cracking can be caused by capillary pore pressure when the pressure gradient is too large. Reducing the amount of microcracking could be implemented by using SRAs to reduce surface tension and slow down the drying process. However, there is much more work to be done in this eld before alkali-activated materials can serve as a full- edged OPC replacement.

Supplementary Information