PSI - Issue 67

E.D. Reis et al. / Procedia Structural Integrity 67 (2025) 39–46 Reis et al. / Structural Integrity Procedia 00 (2024) 000 – 000

41 3

Before producing the concrete samples, the CNTs were pre-dispersed on the cement particles in an isopropanol medium. This method has been validated in previous studies concerning pastes and mortars (Makar and Chan, 2009; Vilela Rocha and Ludvig, 2018) and previously discussed. However, in this research, minor adaptations were made to suit the equipment available in the laboratory. The process involved three distinct steps. Firstly, CNTs and approximately 200 ml of isopropanol were combined in a glass container. Subsequently, the mixture was stirred at 10,000 rpm and subjected to sonication using an ultrasonic apparatus with a frequency of 42 kHz for 30 minutes (Fig. 1a). Following this, 10% of the cement mass and an additional 200 ml of isopropanol were introduced into the mixture. The resulting mixture was transferred to a plastic container, which underwent mechanical stirring and sonication for 90 minutes (Fig. 1b). The final step entailed placing the mixture in a glass container and drying it in an oven at a temperature of 100 ± 5 ºC for 24 hours. This process resulted in a visually uniform dry powder comprising cement particles coated with CNTs (Fig. 1c). The obtained dry mixture was blended with the remaining cement to prepare the samples before adding water.

Fig. 1. CNT dispersion steps: (a) ultrasonication in isopropanol; (b) mechanical stirring of CNTs and cement in isopropanol; (c) drying and mixing with remaining cement.

The mixture consisted of 384 k/m 3 of cement, 960 kg/m 3 of natural sand, 838 kg/m 3 of gneiss gravel, and 204 kg/m 3 of water (w/c ratio of 0.53). Three samples were produced: C0 (without nanotubes), C0.05 (0.05% CNTs added — 0.192 kg/m 3 ), and C0.10 (0.10% CNTs added — 0.384 kg/m 3 ). For all groups, 0.26 kg/m 3 of hydration stabilizer and 0.80 kg/m 3 of superplasticizer additives were used, except for sample C0.10, for which it was necessary to increase the superplasticizer dosage to 0.96 kg/m 3 to ensure good workability. A 120 L concrete mixer was used to prepare the mixture sequentially. Initially, a small amount of water was introduced to moisten the inner surface of the mixer. Subsequently, the aggregates were combined with 60% of the total mixing water, which was processed for three minutes. The cement (with or without CNTs), additives, and the remaining 40% of mixing water were added. The entire composition underwent an additional five minutes of mixing. Concrete consistency was assessed through the slump test (ABNT, 1998). Thirty-six cylindrical specimens measuring 10 × 20 cm (diameter × height) were cast and divided into groups, with 12 dedicated to measuring compressive strength, tensile strength, porosity, and ultrasonic pulse velocity — three for each property. All specimens were cast simultaneously, de-molded after 24 hours, and subjected to curing in a saturated calcium hydroxide solution at 23 ± 2 °C for 28 days (ABNT, 2015). Compressive strength was determined using axial compression according to NBR 5739 (ABNT, 2018b), utilizing the Universal Testing Machine (UTM) — EMIC PC 200 CS Model. Tensile strength was measured through diametral compression, following NBR 7222 (ABNT, 2011), with the same UTM. Porosity evaluation was conducted using the method outlined in NBR 9778-2 (ABNT, 2009), involving an oven and hydrostatic weighing balance. Ultrasonic pulse velocity, indicating material integrity, was determined by the propagation of high-frequency sound waves per NBR 8802 (ABNT, 2019), employing the Pundit Lab(+) Model sensor manufactured by Proceq. Scanning electron microscopy (SEM) was carried out on the broken samples using a Hitachi TM3000 microscope to capture any effect of the CNTs. The impact of the individual factor (CNT content) was examined using Tukey's means contrast test, with a significance level set at 5%. Within this test, each sample is assigned a letter representing its mean value for a specific property: "a" denotes the sample with the highest mean value, "b" corresponds to the second-highest mean value, and so on. When treatments share the same letter, their means are statistically equivalent. The Anderson Darling normality test was employed at a significance level of 5% to validate the results of the Tukey test. A p-value greater than or equal to the significance level suggests a normal distribution condition. Finally, the results obtained at 28 days were compared with those in the literature, taking the same properties as a reference for evaluating the influence of CNT dispersion techniques in concrete.