PSI - Issue 64

Laurena De Brabandere et al. / Procedia Structural Integrity 64 (2024) 97–104 Laurena De Brabandere et al. / Structural Integrity Procedia 00 (2019) 000 – 000

101

5

= 4.04 ∙ ∙ ∙ 2 ∙ ( 12 − 22 ) ∙ 10 −16

(2)

3.4. Torrent permeability The Torrent permeability is measured using the Torrent permeameter. During this test, a vacuum is created inside the inner and outer chamber using a pump. After a certain time, when the desired pressure is reached, the pump is turned off and the pressure increase in the inner chamber is measured over time. This rate is then used by the apparatus to calculate the Torrent gas permeability coefficient k T (Romer, 2005). The Torrent permeability is measured on the same samples that are used for the CEMBUREAU method, immediately after the specimens were taken out of the permeameter cells.

4. Results and discussion 4.1. Capillary imbibition

Fig. 3 shows the capillary water uptake in the specimens as a function of the fourth root of time for specimens with a w/c of 0.4 (Fig. 3a) and 0.7 (Fig. 3b). The linear regression line represents the capillary imbibition rate (CIR). As expected, the CIR for samples with a w/c of 0.7 is larger in comparison to samples with a w/c of 0.4. For a w/c of 0.4, CEM III has the highest CIR and therefore has an inferior performance in comparison to CEM I and CEM II, which have a similar performance. However, this trend is not shown for samples with a w/c of 0.7. Here, CEM III has a superior performance, followed by CEM I and then CEM II which had the worst performance. These results show that the performance of the different cement types in terms of water absorption depends on the w/c of the mixture. The lower CIR for CEM III with a w/c of 0.7 can possibly be explained by a reduced pore connectivity.

Fig. 3. Capillary water uptake as a function of the fourth root of time for specimens with a w/c of (a) 0.4 and (b) 0.7 for CEM I, CEM II and CEM III. Error bars represent the standard error.

4.2. Water vapour diffusion Fig. 4 shows the mass gain per unit area multiplied by the thickness (M/A ∙ Δx) as a function of time. The linear regression line represents the water vapour transmission (WVT) considering the thickness, WVT ∙ Δx. For samples with NaCl as a hygroscopic salt (lowest moisture gradient tested), there is a similar trend compared to the results of the capillary imbibition test. For a w/c of 0.4 (Fig. 4 a), CEM III has a higher WVT ∙ Δx in comparison to CEM I and CEM II, while for a w/c of 0.7 (Fig. 4c), CEM III has a better performance in comparison to CEM I and CEM II. For samples with LiCl as a hygroscopic salt, the difference between CEM I, CEM II and CEM III is smaller for both w/c 0.4 (Fig. 4b) and 0.7 (Fig. 4d). An ANOVA analysis showed that for these samples there was no significant statistical difference between the 3 cement types for the WVT ∙ Δx of w/c 0.4 (level of significance = 5%, p-value = 11.4%) and 0.7 (level of significance = 5%, p-value = 94.8%).