Assessment of physical and mechanical properties of concrete with carbon nanotubes pre-dispersed in cement

The expanding field of nanoengineering modification of cementitious materials increasingly supports advances and innovations in science and engineering through nanotechnology. In this context, this research evaluates the physical and mechanical properties of concrete with carbon nanotubes (CNT–concrete), including compressive strength (CS), tensile strength (TS), static (Es) and dynamic modulus of elasticity (Ed), water absorption (WA), porosity (P), bulk density (ρap), weight loss (WL), ultrasonic pulse velocity (UPV), and electrical resistivity (ER). Specifically, this paper investigates the correlations between these properties and proposes quadratic regression models based on CNT contents to estimate them. For this, an extensive experimental campaign was carried out. As a novelty, CNT were pre-dispersed in the cement particles in an isopropanol to produce the concrete, a proven effective process but used only in pastes and mortars so far. Concrete samples were produced with CNT contents of 0% (as a reference), 0.05%, and 0.10% by weight of cement, with 30 cylindrical specimens measuring 100 × 200 mm (diameter × height) being molded for each composition. The results revealed that including 0.05% CNT led to statistically significant results, such as reductions of up to 12% in P and 14% in WA and increases of up to 16% in CS, 29% in TS, and 3% at UPV. At 0.10% CNT, Ed showed up to 10% improvements, while WL increased to 28%. It was also shown that pre-dispersing powdered CNT in cement particles using an isopropanol medium is effective for use in concrete, but dispersing CNT in water is simpler, safer, and potentially more efficient, especially for industrial production. Another important conclusion was the identification of significant correlations between some properties, which varied with the amount of CNT added, suggesting the existence of a CNT saturation point. Furthermore, the regression models identified ideal CNT contents for P, WA, CS, TS, and UPV, ranging between 0.045% and 0.123% CNT. This research also points to the need for additional studies to improve the quality of the regression models and thoroughly assess the benefits of incorporating CNT into concrete studies.