3.1. Fresh Properties
The self-compacting properties of the concretes were determined according to the standards set by [
28,
29,
30]. The fluidity of the concretes was measured using the slump flow test based on the flow measurements. According to [
28], concretes can be classified by their flow as SF1 (between 550 and 650 mm), SF2 (between 660 and 750 mm), or SF3 (between 760 and 850 mm), with each class having a specific application.
As shown in
Figure 5, the SCLC-EC and SCLC-SBA50 concretes were classified as SF2, suitable for most reinforced concrete applications, while the SCLC-SBA90 and SCLC-SMR37 concretes were classified as SF1, recommended for structures with low reinforcement rates, self-compacting concretes that require pumping, and/or structures that demand little horizontal flow.
The difference in flow between SCLC-EC and the other concretes can be attributed to the low density of the expanded clay aggregate and its vitreous surface, resulting in easier flow of the cementitious matrix due to the rolling action of the aggregates [
17].
Similarly, although the aggregate in SCLC-SBA90 had a higher density than the aggregate in SCLC-SMR37, the greater flow of SCLC-SBA90 can be explained by its water absorption. The [
35] states that ideal water absorption values for lightweight aggregates typically range from 5% to 25%. Thus, while the SBA90 aggregate had a water absorption of 15%, which facilitated better homogenization between the aggregate and the cementitious matrix and was closer to the expanded clay’s absorption of 14%, the SMR37 aggregate had a water absorption rate of 2.5%, which made it more difficult to bond quickly with the cement paste.
In
Figure 6, the results of the flow time test (T500), which measured the apparent viscosity of the concrete, are presented. Apparent viscosity is related to the consistency of the mixture and directly influences the concrete’s resistance to flow.
According to the results, all the concretes were classified as VS2, making them suitable for most common applications. This classification also indicates a thixotropic effect, meaning the concrete exerted less pressure on the forms and had improved resistance to segregation, as per the definitions established by [
28].
In agreement with Barnat-Hunek et al. [
21],
Figure 7 shows a relationship between the increase in fine aggregate consumption and the viscosity of the concretes. As the amount of sand in the mixture increased, the flow time also increased, a result of the density differences between the aggregates and the cementitious matrix [
27,
28].
In Anjos et al. [
29], it was observed that, despite the concretes having the same viscosity classification, mixtures with higher amounts of mineral additions, particularly metakaolin, exhibited higher viscosity values. In the present study, although the amount of metakaolin was the same for all mixes, a directly proportional relationship between the increase in viscosity and the amount of fines in the mixture was noted.
Additionally, Almawla et al. [
30] and Nahhab and Ketab [
23], found an inverse relationship between the volumetric amount of lightweight aggregates and flow time in their studies. However, in the results obtained in this study, the opposite was observed: increasing the amount of aggregates in the mix increased the flow time. This behaviour was consistent across all mixtures, except for SCLC-SMR37, which can be explained by the low absorption rate of the aggregates that prevented complete homogenization with the matrix.
Figure 8 illustrates the behaviour of the concretes during the slump flow test, which provided the visual stability index. This qualitative parameter evaluated concrete’s resistance to segregation and/or bleeding. According to the observations, all the mixtures were classified as VSI 0, as per [
29], meaning that all the mixtures flowed uniformly without evidence of segregation or bleeding, with aggregates visible to the edge of the flow.
Finally, passing ability, which evaluates the concrete’s ability to flow around obstructions, was assessed using the J-ring test.
Figure 9 presents the passing ability results for the SCLCs.
All the concrete mixes were classified as PJ2, meaning they were suitable for most applications and for structural elements with reinforcement spacing between 60 and 80 mm, according to the parameters established by [
1].
Figure 10 shows the performance of each mix during the test, indicating that all mixtures experienced slight aggregate accumulation in the centre, though no signs of segregation or bleeding were observed at the edges.
The last characteristic evaluated in the fresh state was bulk density.
Figure 11 shows the average fresh bulk density values of the concretes studied, along with their respective standard deviations.
Due to the high absorption of the aggregates in the SCLC-SBA50 and SCLC-SBA90 mixes, with absorption rates of 25% and 15%, respectively, the fresh bulk density values exceeded the normative limit for lightweight self-compacting concretes. These higher values were associated with the amount of water absorbed by the aggregates during the saturation period. Additionally, it can be observed that the mixes with the lower volume of lightweight aggregate per cubic meter corresponded to higher bulk density values. These results align with the findings of Nepomuceno et al. [
13], which indicated that reducing the volume of lightweight aggregate led to an increase in concrete density, establishing an inversely proportional relationship between these factors.
Based on the results of the fresh concrete characterization tests, one may conclude that all mixtures met the requirements established in the standards, demonstrating the feasibility of using the new aggregates for producing SCLC when compared to commercial expanded clay aggregates. It is worth noting that the mix with the SBA50 aggregate, despite the differences in characteristics compared to the commercial aggregate, showed consistent performance, with values close to those of the SCLC-EC mix.
3.2. Mechanical Properties
For the hardened-state properties, the results of the mechanical compressive strength test are discussed. The concretes were subjected to axial compressive strength tests at 7 and 28 days.
Figure 12 presents the average compressive strength values and the respective standard deviation ranges for each mix.
As shown in
Figure 12, the SCLC-EC sample exhibited greater variation in compressive strength values compared to the other mixtures studied. This behaviour can be explained by the intrinsic nature of the commercial lightweight aggregates used (expanded clay), which have greater porosity and can present variable mechanical properties, such as crushing resistance, depending on their microstructure and production process. Previous studies [
2,
13] have highlighted that this variability is a characteristic challenge of the use of lightweight aggregates in structural concretes.
Another potential factor is the influence of the surface texture of commercial aggregates. Aggregates with lower uniformity may affect the homogeneity of the cement matrix during mixing, creating localized zones of lower or higher density, which directly reflects on the dispersion of the strength results. In addition, the rheological behaviour of SCLC-EC, combined with the lower water absorption of expanded clay (14%, according to
Table 2), may have reduced the efficiency of the interaction between the aggregate and the cement paste, impacting the paste–aggregate transition.
Despite the greater variability, SCLC-EC presented average strength values that exceeded the minimum limits established by ACI 213 R-03 for structural lightweight concrete. This result reinforces its technical feasibility and highlights the importance of considering the variability of commercial aggregates when designing lightweight concretes. Future studies could focus on the detailed analysis of the aggregate microstructure and the refinement of the mixing process to mitigate these effects.
Based on the results obtained, all concretes, even at 7 days, met the minimum compressive strength requirements for structural concrete as per [
36,
37,
38], which sets a minimum strength of 20 MPa, and for lightweight structural concrete as per [
35], which requires a minimum strength of 17 MPa. At 28 days, the strength values significantly exceeded the minimum required for lightweight concretes, showing considerable strength gains.
In the work of Angelin et al. [
31], it was found that as the amount of larger lightweight aggregate (D
max > 10 mm) increased, the compressive strength decreased, supporting a previous study [
28]. Therefore, a solution to mitigate the influence of larger aggregate diameter is to partially replace the larger lightweight aggregate with a smaller lightweight aggregate (less than 10 mm). However, this may increase the concrete bulk density in most cases [
2,
32].
Since the aggregates used in the mixes had a standardized maximum diameter of 9.5 mm, this behaviour was not observed in the results of the present study. However, the influence of these factors became noticeable only after 7 days. According to Angelin et al. [
31], lightweight aggregates do not affect compressive strength at early ages, as the cementitious matrix is responsible for withstanding mechanical stresses. In this regard, the volume paste-to-aggregate (Vp/Vs) ratio directly affects compressive strength, as a lower ratio (<0.60) results in a larger volume of mortar and a smaller volume of coarse aggregate [
13].
Figure 13 presents a relationship between the volume of lightweight aggregate per cubic meter, the mechanical strength from 7 to 28 days, and the water absorption of the lightweight aggregates.
It can be noticed that the highest compressive strength gains corresponded to mixes with larger volumes of aggregate per cubic meter and lower aggregate water absorption values, as seen in the SCLC-EC and SCLC-SMR37 mixes. The other mixtures had lower strength gains, particularly SCLC-SBA50, which, despite having a higher volume of lightweight aggregate compared to SCLC-SBA90, had a higher water absorption value than SCLC-SBA90. High water absorption within the aggregate can directly influence its crushing strength, making the aggregate more brittle and contributing to the lower performance of the concrete in terms of compressive strength.
Thus, in agreement with Bogas et al. [
2], in addition to the proportion of lightweight aggregates in the mix and their water absorption, other specific characteristics of the aggregates, such as crushing strength, can influence the performance of SCLC. A directly proportional relationship between the compressive strength of the studied concretes and the crushing strength of each aggregate was observed, where the mix with the strongest aggregate (SCLC-EC) showed the highest strength gain, while the mix with the weakest aggregate (SCLC-SBA50) showed the lowest strength gain at 28 days.
Conventionally, in normal-density concrete, compressive strength is primarily related to the water-to-cement ratio (w/c) and the amount of binder in the mix [
27,
33]. However, this relationship cannot be applied effectively to concretes with lightweight aggregates due to the difficulty of determining how much of the mixing water will be absorbed by the aggregates. In this study, the water-to-cement ratio (w/c) was not a significant variable, as the same ratio was maintained in all mixtures. Furthermore, the internal water absorbed by each aggregate during the pre-saturation process was not considered in the calculations, following the premise that this water remains retained within the aggregates and is not released during mixing, as supported by [
35].
Therefore, the lightweight aggregate becomes the main factor influencing not only the strength of the concrete, but also all its other properties [
27]. Supporting the compressive strength results,
Figure 14 shows the self-compacting lightweight concrete specimens after failure, allowing for observation of the aggregates’ behaviour within the cement matrix in terms of dispersion and homogeneity.
It is evident in all the concretes that, unlike conventional concrete, the fracture surface passed through the aggregates, reaffirming the lightweight aggregates’ influence as the limiting factor in compressive strength. According to previous studies [
34,
35], this occurs due to the improvement of the paste–aggregate transition zone and because there is no significant difference between the elastic moduli of the paste and the lightweight aggregate, leaving the lightweight aggregate as the weakest element in the mixture.
Parallel to this, dry bulk density was measured at 7 and 28 days of curing, along with the compressive strength test.
Figure 15 presents the dry bulk density values at 7 and 28 days of curing and the average dry bulk density for each studied mix.
From the results, it is noted that only SCLC-SBA90 showed a bulk density above the limit defined by [
38], which is 2.00 g/cm
3. However, this behaviour can be explained by the higher amount of lightweight aggregate in the SCLC-SBA90 composition, as seen in
Table 5, which led to greater water absorption. One possible solution would be to increase the aggregate size [
23], as larger lightweight aggregates have lower bulk density and lower water absorption [
18]. Moreover, considering that the variation is not significant, previous studies [
20,
23,
31,
36] have found that bulk densities greater than 2.03 g/cm
3 still met the classification for lightweight self-compacting concrete, as the other requirements were fulfilled.
For instance, Bogas et al. [
2] reported compressive strength values between 22.5 MPa and 30 MPa at 28 days for expanded clay-based self-compacting lightweight concrete, with densities ranging from 1.80 to 1.95 g/cm
3. In comparison, our SCLC mixes, such as SCLC-EC and SCLC-SMR37, achieved compressive strengths up to 36.72 MPa and densities between 1.94 and 2.03 g/cm
3, demonstrating strong performance with unconventional aggregates like SBA and SMR.
Additionally, studies by Santis and Rossignolo [
24] and Lotfy et al. [
32] reported similar results for expanded clay and pumice-based lightweight concretes, with compressive strengths around 30 MPa and densities of approximately 1.85 g/cm
3. Our SCLC mixes not only matched but exceeded these values, reinforcing the potential of SBA and SMR as viable alternatives for producing high-performance self-compacting lightweight concrete.
Finally, the efficiency factor (EF) was evaluated, which is the ratio between compressive strength and concrete bulk density. In this study, the efficiency factor was calculated for the concretes at 28 days, using the average dry bulk density value.
Figure 16 presents the efficiency factor for each SCLC mix.
Based on the results, it can be observed that the efficiency factor for the self-compacting lightweight concretes ranged from 13 to 18 MPa·dm
3/kg, with the highest values found in SCLC-SMR37 and SCLC-EC, which were the mixes that exhibited the highest compressive strength and lowest bulk density. Previous studies have found efficiency factors ranging from 22 to 28 MPa·dm
3/kg [
31]. However, in both studies, different aggregate sizes were used, in addition to partial replacement of lightweight aggregate with normal coarse aggregate, which increased the compressive strength of the concretes and, consequently, the efficiency factor.
According to Santis and Rossignolo [
24], the efficiency factor is an important parameter in structural design, and concretes with efficiency factors above 25 MPa·dm
3/kg can be considered high-performance concretes. In this study, all the mixes were classified as structural concretes, meeting the compressive strength and bulk density parameters defined by [
35].
3.3. Durability Properties
To determine the chloride ion diffusion coefficient in the studied concretes, the procedures established by [
34] were followed. In this test, chloride ion penetration occurred in a single direction, from bottom to top, driven by the passage of electric current.
Figure 17 shows the concrete specimens after the test and the application of silver nitrate (AgNO
3) spray to define the chloride ion penetration depth.
Qualitative evaluation showed that the SCLC-SMR37 mix exhibited the smallest chloride ion penetration front, followed by SCLC-EC, SCLC-SBA50, and finally SCLC-SBA90, which showed the greatest penetration depth. However, it was noted that the aggregates in the SCLC-SMR37 mix floated within the paste and deposited on the top part of the specimen, resulting in segregation between the cementitious matrix and the aggregates. This behaviour can be explained by the low absorption and relatively low density of the SMR37 aggregate, making it difficult to homogenize with the paste as it requires more time to absorb the mixture.
Due to the behaviour exhibited by the SMR37 aggregate, in accordance with the normative specifications of [
34], the chloride ion diffusion coefficient for SCLC-SMR37 was disregarded as it did not represent the proper behaviour of the mix.
Figure 18 presents the chloride ion diffusion coefficients for the self-compacting lightweight concretes at 186 days, after being subjected to wet and air curing.
From the obtained results, a relationship between the diffusion coefficient and the mortar content of each mix can be observed, as per the data in
Table 5. It is noted that the mixes with lower diffusion coefficients, except for SCLC-SMR37, had higher mortar contents. According to Gonen and Yazicioglu [
19] and Nahhab and Ketab [
23], the presence of porous aggregates contributes to increased permeability in concretes. However, a higher mortar content produces a denser cementitious matrix, reducing permeability by improving the matrix’s pore structure, thus enhancing durability performance.
Additionally, there is microstructural improvement due to internal curing provided by the water retained by the lightweight aggregate, which refines the aggregate–matrix interface, reducing microcracks typically found in concretes with conventional aggregates. These microcracks are the main pathways for aggressive agents to percolate.
Real et al. [
37] reported that lightweight aggregate itself does not influence the determination of the diffusion coefficient, indicating no clear relationship between aggregate porosity and the diffusion coefficient. However, other factors such as matrix quality, w/c ratio, binder content, and especially pre-saturation of high-absorption aggregates can affect the concrete’s resistance to chloride ion penetration. It is recommended to avoid saturating the aggregates to reduce their contribution to chloride penetration.
As observed in the study and discussed by other authors [
19,
23,
38], this may have occurred due to the low mortar content in the mixture, as well as the higher proportion ratio between coarse aggregate and mortar (Vg/Vm). Alternatively, as Real et al. [
37] found in their study, although lightweight aggregate alone does not influence the determination of the diffusion coefficient, lightweight aggregates with high absorption levels can influence the resistance of concrete to the entry of chloride ions. However, when comparing CLAA-RBC50 (25% absorption) with CLAA-RBC90 (15% absorption), we observed that the diffusion coefficient of CLAA-RBC50 was lower, even though it had the aggregate with the highest absorption. However, when compared to CLAA-RBC90, CLAA-RBC50 has a higher mortar content and a lower Vg/Vm ratio, confirming what was stated in previous studies: The higher mortar content produces a denser cementitious matrix, contributing to the reduction in concrete permeation through the improvement of the matrix pore structure, thus contributing to better performance in terms of the durability of the structures.
Considering this, the results obtained allowed for a comparison between the volume of lightweight aggregate per cubic meter and resistance to chloride ion penetration, as shown in
Figure 19.
It is evident that, except for the SCLC-SMR37 mix, the higher the volume of lightweight aggregates in the mix, the lower the chloride ion diffusion coefficient. The porosity of the aggregates does not have a significant direct relation with the resistance to chloride ion penetration. Thus, it can be concluded that matrix quality plays a larger role than lightweight aggregates in influencing durability, as the mix with the greatest susceptibility to chloride ion penetration corresponds to the one with the highest aggregate-to-mortar ratio (Vg/Vm) and the lowest mortar content.
Table 7 presents a summary of the results obtained in this study.
SCLC-SMR37 presented the best overall performance, particularly in compressive strength, density, and resistance to chloride ion diffusion. SCLC-SBA90 presented competitive results in some parameters, such as compressive strength and apparent density. However, it was disregarded in the chloride ion diffusion analysis due to less satisfactory behaviour compared to SCLC-EC. Based on the overall balance of performance metrics, SCLC-SMR37 is indicated as the most efficient blend.
The effect of aggregates on the properties of self-compacting concrete (SCC) has been extensively studied in the literature, with numerous international studies emphasizing the significant impact of lightweight aggregates on workability, strength, and durability. For example, Gopi et al. [
8] emonstrated that the use of expanded clay aggregates in SCC notably enhanced its workability, achieving slump flow values comparable to those observed in the present study for SCLC-EC. Similarly, Ting et al. [
28] highlighted that aggregates with lower density, such as those derived from sugarcane bagasse ash (SBA), improve the flowability and passing ability of SCC, aligning well with the results for SCLC-SBA50 and SCLC-SBA90 in this study.
Further studies by Bogas et al. [
2] and Lotfy et al. [
21,
32] underscore the critical role of water absorption characteristics in lightweight aggregates, which directly influence viscosity and compressive strength. In this study, SCLC-SBA90, with its higher water absorption rate, exhibited a moderate increase in flow time (T500) compared to SCLC-EC. This observation is in line with findings by Barnat-Hunek et al. [
21], who reported that aggregates with higher water absorption led to increased viscosity, which in turn affected the overall workability of the concrete.
When comparing the compressive strength results, this study’s findings are consistent with those of Santis and Rossignolo [
24], who found that mixes with smaller lightweight aggregates demonstrated higher compressive strength gains due to a better interfacial bond between the aggregate and the cement paste. Our study, SCLC-SMR37, which showed the highest strength at 28 days, follows this trend, further emphasizing the importance of aggregate properties in enhancing concrete strength.
Regarding chloride ion diffusion, the research by Real et al. [
37] and Nahhab and Ketab [
23] highlights that aggregates with lower porosity and water absorption rates significantly improve resistance to chloride ion penetration. This was evident in the performance of SCLC-SMR37 in this study, which demonstrated a lower chloride diffusion rate compared to other mixtures.
Overall, the results of the present study align well with international research, reinforcing the potential of using unconventional aggregates such as SBA and SMR for producing sustainable self-compacting lightweight concrete. These aggregates not only offer environmental benefits, but also demonstrate optimized performance in both the fresh and hardened states.