3.2.1. Water Absorption through Capillarity
The characterization of all mixtures had already been carried out in mortar matrices [
48], which is why not all the results are presented, but only the final selected mixtures with fly ash addition. The results revealed that mixtures with higher packing density, both BT and MT, and combined with slag addition, presented the best results, absorbing down to 60% less water. Fly ash and pozzolan addition also revealed good performance in reducing water absorption.
Figure 9 shows the evolution of capillary absorption values, S
a (mg/mm
2) with the square root of time, for BT and MT mixtures, after 28 days curing. By analyzing the results, performed on three consecutive days, it can be concluded that the capillary water absorption occurs with more intensity in the first hours and tends to stabilize over time, presenting a nonlinear evolution. In both series, the decrease of the powder content combined with the increase of compactness and the partial replacement of cement by additions proved to be advantageous in decreasing water absorption over time. However, this fact is more prominent in MT series, since the values of optimized mixtures in BT series remain similar to the reference. In the latter, the addition of fly ash (50% of replacement rate) and the inclusion of 20% of recycled aggregates does not change the capillary absorption of concrete. Comparing the two series, it is noticeable that the amount of cement influences this parameter, proved by the lower values of the BTref (with cement content of 400 kg/m
3), in relation to MTref (with 400 kg/m
3). However, the optimized mixtures with fly ash show that the increase in compactness and the replacement of cement by this addition also has a great influence on the reduction of capillarity, as the concrete matrix becomes more compact and less permeable. In BT series, despite of the 50% of cement replacement, the capillary absorption is similar to reference. However, in MT mixtures, the optimized matrix with high compactness, and about 30% of cement replacement by fly ash, promotes a significant reduction in capillarity, of at least 35%.
In MT mixtures, the partial cement replacement by fly ash and 20% of recycled concrete aggregates, mixture MT250-FA-RA, leads to down to 44% less water absorption, and the mixture with natural aggregates (MT250-FA) presents equally reduced values.
Figure 10 shows the quality of the concrete as a function of the capillary absorption coefficient, S
a. Browne [
69] proposed the following classification for the quality of the concrete, as a function of the S
a coefficient: above 0.2 mg/mm
2 × min
1/2 is “low quality”; between 0.1 and 0.2 mg/mm
2 × min
1/2 is “medium quality” and below 0.1 mg/mm
2 × min
1/2 is “high quality”. According to this classification, all mixtures are considered to be of “high quality”, regarding the water absorption through capillarity. In addition, due to pozzolanic effect of the optimized mixtures, there is great potential to improve the performance for longer ages, beyond 28 days.
3.2.3. Carbonation Resistance
The phenomenon of concrete carbonation is mainly associated with the penetration, by diffusion, of carbon dioxide. The decrease in pH is visible through the reaction with phenolphthalein sprayed in concrete, in which the carbonated area is colorless (
Figure 12).
Figure 13 shows the results evolution of the carbonation depth (C
di), measured in mm, of the BT and MT series, with the square root of time. The analysis of carbonation depth, C
depth, of the studied mixtures proved to evaluate that it varies with the square root of time, as expected, with a general trend to be linear. The amount of cement highly influences this parameter, and that effect is more evident in BT mixtures, whose reference contains a higher amount of cement than in MT mixture. However, BT250-slag mixture proves that the slags are the most beneficial additions in this series, showing a slower initial evolution and only 46% of increased carbonation depth in relation to the reference. On the contrary, also in BT series, the increased compactness and the partial replacement of cement by fly ash (BT200-FA and BT200-FA-RA) is not able to avoid the increase of the carbonation depth, up to 122%, when compared to the reference. Through these mixtures, it can also be concluded that the inclusion of 20% of recycled aggregates does not increase carbonation, on the contrary, it is slightly lower. Previously to this study, in mortar matrices, pozzolans additions did not reveal promising results in preventing carbonation [
48]. In concrete mixtures, the results proved to follow the same trend.
In the MT series, the difference between the optimized mixtures and the reference is not so evident, since the cement replacement is lower (30%), presenting a similar evolution of the carbonation depth for the first testing ages. This evolution is less pronounced after 56 days, with the exception of the MT250-slag mixture, which shows a more pronounced increase of the carbonation depth, presenting values up to 69% above reference. This effect was not noticed for younger testing ages on the preliminary study of mortar matrices, since the carbonation evolved linearly with square root of time up to 90 days. Unlike the BT series, the optimized mixtures with fly ash of MT series, MT250-FA, present values closer to the reference, showing an increase of up to 31%, at the longer tested age. The inclusion of recycled aggregates, once again, did not harm the performance of the concrete at this parameter, presenting similar evolution with lower values to the homonymous mixture with natural aggregates. Despite additional absorption, water is considered in concrete with RA, the probable absorption during setting time may reduce the effective water of the mixtures with RA, which seems to result on a positive effect on carbonation reduction.
3.2.4. Chloride Migration
The penetration depth of chlorides into the specimen is visible through the silver colour area, which corresponds to the silver nitrate precipitation, as is shown in
Figure 14.
The results shown in
Figure 15 represent the chloride diffusion coefficient (D
0) by non-steady state migration test at 56 and 90 days. In this test, the lower the D
0 coefficient, the higher the resistance to chloride migration, the durability performance also being higher. This test highlights the benefit of pozzolanic additions (both fly ash and pozzolans), bringing the pozzolanic refinement to the porous network of the matrix, combined with increased compactness, in reducing the chloride migration.
The addition of natural pozzolans stands out positively by presenting very low D values, down to 64% and 81%, below the references for mixtures BT200-Poz and MT250-Poz, respectively, at 90 days. Due to its high content of alumina, when pozzolans are combined with cement, it promotes a reduction of the porous structure of the matrix with time, being beneficial to durability related with to corrosion risk, induced by chloride penetration, and to strength performance.
The results reveal similarly low values for the BT200-FA and MT250-FA mixtures, down to 80% and 77%, respectively, at 90 days, when compared to the references. This effect is already scientifically known. However, the inclusion of recycled aggregates affects the increase in the diffusion coefficient, when compared with similar mixtures with only natural aggregates, although it presents values 51% and 70% below the references for BT200-FA-RA and MT250-FA-RA, respectively. It is important to note that, in the BT200-FA-RA mixture, the D0 coefficient maintain its value from 56 to 90 days. Contrarily, the MT250-FA-RA mixture reduces the diffusion coefficient. Apparently, the higher cement content and lower replacement rate, in comparison to BT mixtures, is the probable reason for that difference.
Figure 16 show the classification of the chloride penetration resistance of concrete as a function of the chloride diffusion coefficient, D
0, as used in [
33].
According to this classification, and by analyzing the plotted graph, it is possible to note that only the references, both BT and MT, present a “moderate” resistance of concrete to chloride penetration, although the BTref mixture achieves a classification of “high” resistance, at 90 days. Mixtures optimized with additions and increased compactness reveal very promising classification results. The mixtures with the addition of pozzolans (more evident in MT250-Poz) stand out for mostly obtaining the classification of “extremely high” resistance to chloride penetration. Mixtures with fly ash also stand out for the same reason, although they are mostly classified as “very high” resistance. The replacement of natural aggregates by recycled concrete aggregates does not greatly harm the mixtures, since, after 90 days, they are maintained, at least, with the classification of “very high” resistance.
3.2.5. Concrete Surface Resistivity
Figure 17 shows the results of the surface electrical resistivity of concrete (ρ), of the BT and MT mixtures, over time. This is a fast test allowing to indirectly predict the permeability to fluids and the diffusion of chloride ions in concrete. The presented results were determined in mortar matrixes of the same concrete formulations, since they show the same trend as in concrete.
Through the analysis of the graphs, in both series, it is possible to conclude that the slag addition does not lead to positive results in this test, presenting similar values to the references with plain cement binder. The reference mixtures and those with slag addition have no significant variations of electrical resistivity with age. The opposite is found in mixtures with pozzolanic additions, which present results that increase with the testing age. Mixtures with fly ash addition stand out for having higher values, almost 20 times higher than the references, for both BT200-FA and MT250-FA mixtures, measured at 450 days. The pozzolan addition shows resistivity values nearly 6 and 16 times higher than the references, for BT250-Poz and MT250-Poz, respectively, also at 450 days. Despite fly ash and natural pozzolan both being pozzolanic additions, their influence on resistivity is different with time regarding the evolution and the optimal replacement rate. The electrical resistivity evolves faster with age, up to 56 days, for mixtures with pozzolan addition and develops a smoother evolution after that. Contrarily, the mixtures with fly ash addition present smooth evolution at younger ages, but after 120 days the increase in rate with age tends to be faster, probably due to higher pozzolanic reactivity at greater ages. However, at 300 days, mixtures with both additions tend to stabilize, although mixtures with fly ash stabilize more smoothly. In the MT series, the MT250-Poz mixture is very close to MT250-FA, showing closer values of resistivity evolution over time; a lower cement replacement by pozzolans, in comparison to the BT series, seems to suggest a possible optimal for moderate replacement rates (circa 30%). Contrarily, in the BT series, the higher cement replacement of fly ash in the mixture BT200-FA, 50%, suggest that the resistivity increase is more effective for higher fly ash content, mainly for longer ages. In fact, in a previous study carried out in mortar matrixes [
48], it was possible to draw this conclusion, corroborated also by Kurda et al. [
70]. The optimal proportions, for both types of pozzolanic additions, regarding the increase in electrical resistivity and the reduction of chloride migration need to be studied in future research. Further, due to its known relation to the resistance to chloride migration, and as is evident from the analysis of the graphs, the pozzolanic effect promotes an improved behavior at older ages. That effect increases the refinement of the internal porous microstructure with age, both for pozzolan and for fly ash; however, at different rates, the fly ash being lower at shorter ages, but with a greater effect in longer ages, and mainly when high replacement rates are used.
The followed standard [
64] refers to a correlation between the resistivity test and the chloride ion penetration, establishing limit parameters. At 450 days of age, the mixtures with additions of pozzolan and fly ash of the BT and MT series were classified, regarding the penetration of chloride ion, as being "very low", with ρ values between 37 and 254 kΩ-cm. For the reference mixtures, at the same age, BTref was classified as “moderate” (12 kΩ-cm < ρ < 21 kΩ-cm) and MTref as “high” (ρ <12 kΩ-cm). Mixtures with slag additions were classified as “high”, which is why concrete mixtures with that addition are not recommended for exposure to chlorides.
Figure 18 proves the high correlation between the chloride diffusion coefficient, D
0, of the mixtures and the surface resistivity, ρ, of the corresponding matrixes, which is very reliable, and was proved by the high R
2 values of the power function.
3.2.6. Prediction of Service Life
The determination of carbonation resistance, R
C65 (kg × year/m
5), of each developed concrete was calculated using Equation (3) and the carbonation depth, C
di. These values were measured after the concrete was exposed for a certain time to a highly concentrated carbon dioxide environment, C
acel, 90 × 10
−3 kg/m
3.
where t
i is the time of exposure in years, and X
i has the same meaning as C
di (carbonation depth).
Table 3 presents the average values of carbonation resistance, R
C65, at 56, 90, and 120 days to minimize the errors related with the measurement of the carbonation depth.
The deterioration of reinforced concrete due to steel corrosion can be divided into two periods: the initiation, tic, and the propagation of corrosion, tp. The minimum propagation period, tp, depends of several factors, such as the exposure class.
Table 4 presents the propagation period, t
p, according to LNEC E-465 [
71]. Thus, considering that the intended service life, t
g, is equal to 50 years, and considering that the electrical poles can be associated to the reliability class (RC1), where the safety factor of the service life is equal to 2.0 [
71], the design period of initiation, t
ic, can be determined using Equation (4) (
Table 4).
The minimum concrete cover required to guarantee the resistance against the reinforcement corrosion due to carbonation, c
min,dur, is determined using Equation (5), considering the R
C65 already calculated for each concrete and t is equal to the design period of initiation, t
ic (
Table 4 and
Table 5).
The meanings of the symbols are: k0 is a factor related with test conditions and is equal to 3, k1 is a factor related with relative humidity, k2 is a factor related with concrete cure conditions and is also 1, n is a factor related with the influence of wetting/drying cycles over time, and t0 is the reference period and is 1 year.
The Equation (5) can also be used to predict the service life, t
g, when the cover used is already predefined. For this purpose, X must be equal to the cover predefined, and t
ic is the unknown variable that should be calculated using the Equation (5). Usually, the randomness of the factors that affect the service life happens during the initiation period, so after the determination of t
ic, the service life, t
g, is calculated adding to this value the propagation period, t
p, presented in
Table 4.
The service life was computed for different exposure classes considering covers equal to 15, 20, and 30 mm, see
Figure 19 and
Table 6. The presented covers only concern durability and were not determined to ensure proper fire resistance nor adequate transmission of bond forces between rebars and concrete. For the standard for concrete poles, the minimum cover is related to the exposure conditions and also to the concrete strength; parameters of composition were not considered.
It is clear that higher cement content provides a better performance regarding durability related with carbonation, since the two reference concrete mixtures had the highest service life. The results of
Table 6 show that the service life expected for the poles, in most cases, was higher than 50 years, expect in environment XC4 (wet regime), meaning that the significant reduction in the amount of cement of the eco-efficient concrete can be compensated by using additions. The fly ash effect is similar in both types of concrete, the BT-dry consistency and the MT-plastic consistency, and the incorporation of the recycled aggregates does not influence the concrete performance. However, the use of ground electric furnace slag is more beneficial with BT250-slag than MT250-slag.
For environments exposed to chlorides, the minimum propagation period, t
p, also depends of several factors, and, according to LNEC E-465 [
71], those values are 0 years for XS1 and XS3, and 40 years for exposure class XS2 [
71]. The chloride diffusion coefficient, D (m
2/s), is determined for each developed concrete using Equation (6) and using the experimentally determined values of non-steady-state migration coefficients, D
0 (
Figure 15). Using Equations (6) and (7) [
71], and defining the cover thickness, X, it is possible to determine the design period of initiation of poles made with each type of concrete and after the service life, t
g. In addition, using the same equations and defining the design period of initiation, t
ic, it is possible to determine the minimum concrete cover required to guarantee the resistance against the reinforcement corrosion induced by chlorides, c
min,dur. For a service life equal to 50 years and considering the already mentioned reliability class (RC1), the design period of initiation, t
ic, can be determined using Equation (4) and is 100 years for XS1 and XS3, and 20 years for exposure class XS2.
where k is a factor that takes into account the curing conditions, the relative humidity and the temperature; n is a factor that considers the chloride diffusion decrease over time; t
0 is the time of the experimental tests to evaluate D
0; t is the exposure time in days; ξ is a parameter related with the concentration of chlorides in the binding paste; X is the cover. The minimum concrete cover required to protect the steel reinforcement against corrosion induced by chlorides, c
min,dur, is presented
Table 7 and the prediction of service life, t
g, for 20 and 30 mm is presented in
Figure 20 and
Table 8.
It is highly improbable that electric poles will be placed in tidal splash or spray zones or will be permanently submerged, but the results show that poles manufactured with these concretes should not be used in such zones. In environments with air-based sea salts, the reference concretes have a performance that is clearly worse than concretes with additions, which is reflected in the values of minimum cover, as well as in the service life expected. Among the concretes with additions, the behavior is relatively similar, showing that the pozzolanic additions can replace the cement dosage in the context of protection against corrosion induced by chlorides. It is also noted that, in this case, the incorporation of recycled aggregates slightly increases the minimum cover and decreases the expected service life. To assure a service life of 50 years for poles manufactured using these eco-efficient concretes, for both types of consistency, the cover requires ranges between 15 and 31 mm, which are the values usually used in the production of this type of structures.