*2.3. Experimental Methods*

The reactivity of the paste with different LCDBA substitution levels was investigated through XRD and TG analysis. GGBFS blended cement pastes were cured at 20 °C and 90 ± 2% relative humidity for a predetermined curing duration (3, 7, and 28 days). Thereafter, the samples were ground and immersed in acetone to stop hydration and were suction filtered using an aspirator. The crushed samples that stopped hydration were ground further and powdered to particles smaller than 106 μm for the XRD and TG measurements. XRD was conducted using a D/MAX 2500V/PC (Rigaku, Tokyo, Japan) with a scan range of 5◦–65◦ 2θ. TG was conducted using a NETZSCH STA 409 C/CD (NETZSCH, Selb, Bavaria, Germany) with a heating rate of 5 °C/min in the 20–1000 °C range.

The fresh properties of GGBFS blended cement concrete were measured using the slump, bleeding, and setting time. The slump of fresh concrete were measured in accordance with ASTM C 94 [20]. The bleeding test of fresh concrete was measured in accordance with ASTM C 232 [21] by drawing off the bleed water until cessation of bleeding. The initial and final setting times were measured using a penetration resistance apparatus in accordance with ASTM C 803 [22]. The measurements were conducted on sieved mortar samples from the mixed concrete. The fresh concrete was cast into cylindrical molds (100 × 200 mm) for the compression test. Following 24 h of air curing, these cylindrical samples were demolded and immersed into water at a temperature of 20 ± 1 °C for additional curing. Compressive strength tests were carried out in accordance with ASTM C 39 [23]. The strength was determined at 3, 7, 28, and 91 days of curing by averaging the tested values of the three replicates.

#### **3. Test Results and Discussions**

#### *3.1. XRD Analysis*

The diffraction patterns of the hardened pastes were analyzed by characterizing the crystal phases in order to investigate the applicability of LCDBA with high alkali content for stimulating GGBFS. The activation effect of LCDBA for GGBFS can be demonstrated by investigating the generation of hydration products for different curing ages. Figure 4 shows the results of an XRD analysis of the pastes, according to the LCDBA substitution level.

As can be seen in Figure 3, all specimens included akermanite (Ca2Mg(Si2O7)) with strong peaks in the XRD analysis. It has been reported that akermanite exists in a crystalline form in raw materials and hardened GGBFS pastes [24,25] and that it can be identified more clearly for cases with LCDBA. In particular, a strong akermanite peak was observed for the 5% LCDBA series at 28 days of curing, indicating that LCDBA affects the hydration of GGBFS in the long term.

**Figure 4.** XRD results of the blended pastes at different curing ages: (**a**) 3 days, (**b**) 7 days, and (**c**) 28 days.

In addition, the strong alkali compound of the cement hydration product—calcium hydroxide (Ca(OH)2)—is known to play a role in promoting the activation reaction of GGBFS [2]. This effect is evidenced by the XRD pattern at 3 and 28 days. For example, it can be clearly seen in the LCDBA-0% series that the higher peak of the Ca(OH)2 at 3 days decreases as the curing age increases through the consuming necessary for the activation of the GGBFS to form calcium-silicate-hydrate (C-S-H) gel. The pastes of the two series containing LCDBA also had high Ca(OH)2 at their early age and the amount decreased as the age increased. However, the decrease of Ca(OH)2 was smaller than that of LCDBA-0% series, although the C-S-H gel peak increased significantly. This phenomenon was the likely cause of the high alkaline LCDBA activator. In the section where 2θ is around 9◦, it is known that ettringite peaks result from the hydration of cement [26]. It should be noted that the ettringite peak was less affected by the incorporation of LCDBA at 3 days, increasing with the higher LCDBA amount at 28 days. This implies that the production of ettringite is increased by the supply of SO3, as with the high SO3 content in LCDBA, at 47.1%. In addition, Mohammed and Safiullah [27] revealed that the amount of ettringite formation highly correlates with the amount of SO3.

#### *3.2. TG Analysis*

The degrees of hydration of GGBFS blended cement pastes were evaluated using the thermogravimetry method and the results are presented in Figure 5. The weight loss at around 100 °C (refer to Section I afterward) can mostly be attributed to the decomposition of ettringite and C-S-H gel [28]. The weight loss at around 450 °C (refer to Section II afterward) can primarily be attributed to the decomposition of Ca(OH)2 to CaO [29].

From the results for Sections I and II, it can be seen that the weight losses of the LCDBA-3% and LCDBA-5% series are more pronounced than those of the LCDBA-0% series for all curing ages. Therefore, the large mass change in Section I implies that a large amount of C-S-H gel initially forms due to the incorporation of LCDBA that contains abundant alkali compounds that promote the reactivity of GGBFS. The large mass change in Section II can be explained that the cement hydration product, Ca(OH)2, was relatively less consumed in the LCDBA-3% and LCDBA-5% mixtures due to the additional supply of alkaline compounds by LCDBA. In the meantime, it should be noted that the differences in mass change between the pastes with LCDBA and without LCDBA at 28 days were relatively large in Section I in comparison to those in Section II. This might be attributed to the fact that, as curing age increases, the alkali components supplied for the pozzolanic reaction of GGBFS in the LCDBA-3% and LCDBA-5% series were also consumed, resulting in a Ca(OH)2 decrease and reaching similar remaining levels of Ca(OH)2 as those of the LCDBA-0% series at 28 days of curing.

(**c**)

**Figure 5.** TG curves of the blended pastes at different curing ages: (**a**) 3 days, (**b**) 7 days, and (**c**) 28 days.

#### *3.3. Workability*

The slump test is widely used and the most well-known test to assess the workability of concrete. In this study, the effect of the LCDBA usage on workability was evaluated using the slump test. Figure 6 shows the slump values and the superplasticizer dosage of the blended concrete, indicating that all

tested series met the target slump level of 180 ± 25 mm. For all concrete series without LCDBA, the superplasticizer amount was fixed at 0.7%. The figure shows that as the water/binder ratio increases the slump increases by 10 mm per 0.05 of water/binder ratio. The blended concrete that contained 3% and 5% LCDBA tended to decrease in slump even though the dosage of superplasticizer increased. For mixtures with 5% LCDBA, all slump results were the same, 160 mm, and the corresponding superplasticizer amounts were 1.0%, 0.9%, and 0.8%, for the water/binder ratios of 0.35, 0.40, and 0.45, respectively. Therefore, the increase in the required superplasticizer amount for maintaining a similar level of workability implies that the porous LCDBA absorbs the mix water during the mixing and thus degrades the workability of the blended concrete.

**Figure 6.** Slump of the blended concrete with different water/binder ratios and LCDBA amount (solid markers and hollow markers stand for slump values and dosage of superplasticizer, respectively).

#### *3.4. Bleeding*

The bleeding of concrete, a necessary part of the life of concrete, was considered to occur when the mix water would raise to the surface of freshly placed concrete. It has been reported that the bleed rate and capacity of GGBFS blended cement concrete highly depends on the GGBFS replacement level and the water/binder ratio [30]. Figure 7 shows the development of bleeding over elapsed time of concrete for the water/binder ratio of 0.40, as a function of the LCDBA replacement level. The ending point and the slope of the curves indicate the bleeding capacity of the mixtures and the bleeding rate, respectively. It should be pointed out from the figure that the bleeding capacity decreases as the LCDBA replacement level increases. In particular, the mixture without LCDBA (the black curve) showed the highest bleeding rate for the first four hours and then it quickly reached the highest value. On the other hand, the mixtures with LCDBA (the blue and red curves) showed that as the LCDBA replacement level increases the bleeding rate becomes slower for the first three hours. After the first three hours, all mixtures showed similar bleeding rates until reach the bleeding capacity. The delayed endings of the bleeding were observed for both mixtures with LCDBA in comparison to the one without LCDBA. This can be attributed to the fact that the porous LCDBA absorbed excess water during mixing because of its high specific surface area, which reduced the initial bleeding rate. In addition, the bleeding time increased as the water absorbed by the LCDBA was slowly released to the fresh mixture.

**Figure 7.** Effect of the LCDBA replacement level on bleeding (W/B = 0.40).

#### *3.5. Setting Time*

The effect of the LCDBA replacement level on the initial and final setting times of GGBFS blended cement concrete is shown in Figure 8. The setting times were defined when the penetration resistance reached the pre-defined criteria in accordance with ASTM C 803 [22]. The initial setting time of the mixture without LCDBA was 340 min but the mixtures incorporating LCDBA were delayed for more than 80 min. The delay in the initial setting time can be explained by the delay in the bleeding. Similar to the results of the initial setting time, the mixture without LCDBA showed the fastest final setting time, at 483 min, and 20- and 40-min delays were observed for the mixtures with 3% and 5% LCDBA replacement, respectively. It should be pointed out, however, that the differences of final setting time between the mixtures with and without LCDBA were smaller than those of the initial setting time. This can be attributed to the activation of GGBFS promoted by the alkali supply of LCDBA after initial setting.

**Figure 8.** Effect of the LCDBA replacement level on penetration resistance (W/B = 0.40).

#### *3.6. Compressive Strength*

The compressive strength development of GGBFS blended cement concrete with different LCDBA replacement levels and with water/binder ratios at 3, 7, 28, and 91 days is shown in Figure 9. The concrete mixtures that incorporate LCDBA showed a generally higher strength than the mixture without LCDBA. In particular, the incorporation of LCDBA was more effective for early strength development and for mixtures with lower water/binder ratios.

**Figure 9.** Effect of the LCDBA replacement level on compressive strength.

The rates of increase in the compressive strength of the concrete incorporating LCDBA relative to the strength of concrete without LCDBA were evaluated for each curing age, with the results summarized in Table 3. At 3 days of curing, the strength improvement rates of the concrete with 3% and 5% replacement were 13.1–16.7% and 15.1–22.0%, respectively. On the other hand, the strength improvement rates of the concrete with 3% and 5% replacement at 91 days of curing were 1.4–5.8% and 4.1–8.8%, respectively. Therefore, as the replacement level of LCDBA increased, the strength increase rate became higher, while the strength increase rate gradually decreased as the curing age increased. This tendency is similar to the typical pattern of rapid strength development of GGBFS promoted by alkaline activators (see also [31–33]). It should be pointed out, therefore, that LCDBA could be a sustainable and alternative activator for GGBFS blended cement concrete, with similar effects as those of conventional alkaline activators, which is effective for the early-age strength development of concrete.

**Table 3.** Rate of increase in the compressive strength of concrete incorporating LCDBA.


In order to quantitatively analyze the strength developing characteristics of GGBFS blended cement concrete incorporating LCDBA, the compressive strength prediction model suggested in ACI 209 was used. The ACI Committee 209 [34] recommends the following equation for predicting the compressive strength of concrete with time

$$(f'\_c)\_t = \frac{t}{a+b \times t} (f'\_c)\_{28'} \tag{1}$$

where *a* and *b* are material constants considering the type of binders and curing methods, calculated in this research based on regression analysis using measured values. In addition, (*f <sup>c</sup>* )*<sup>t</sup>* and (*f <sup>c</sup>* )<sup>28</sup> are compressive strength at the age of *t* and 28 days, respectively.

The concrete mixtures with 5% LCDBA replacement and without LCDBA were considered to quantitatively illustrate the effect of LCDBA on the development of the compressive strength of concrete. Figure 10 shows the compressive strength results, according to various curing ages and water/binder ratios for the concrete mixtures with 0% or 5% of LCDBA and with the regression curves based on the ACI 209 model.

**Figure 10.** Comparison between the measured compressive strength and the regression curves for 0% or 5% LCDBA replacement series; (**a**) LCDBA-0%, (**b**) LCDBA-5%.

As a result of the regression analysis of the compressive strength, according to the ACI 209 model, the R-squared values for the mixtures with and without LCDBA were higher than 0.985, indicating a high goodness of fit. With the high goodness of fit, the material constants *a* and *b* were used to further analyze the strength development patterns of the tested concrete. The calculated material constants *a* and *b* are illustrated in Figure 11, where the constants *a* and *b* considerably correlate with the strength development of concrete at early age and at long-term age, respectively [35]. Mathematically, *a* is inversely proportional to the initial strength development and *b* is inversely proportional to the increase in compressive strength according to the curing age. As shown in the figure, *a* and *b* tend to decrease and increase, respectively, as the water/binder ratio increases. As the water/binder ratio decreases, the initial strength development rate becomes high, which leads to a decrease in *a*. On the other hand, the material constant *b* is related to the long-term strength development, where the lower *b* value stands for the higher long-term strength increment. The constant *b* values for the mixtures with LCDBA were similar, regardless of the water/binder ratio, and higher than those for the mixtures without LCDBA. The higher *b* values for the mixture with LCDBA were attributed to the higher initial strength development activated by LCDBA and to relatively less long-term strength development. This is similar to the strength development characteristics of GGBFS and/or fly ash blended cement concrete activated using conventional alkali activators. Therefore, it can be concluded that LCDBA is an effective and sustainable alternative to alkali activators for GGBFS blended cement concrete.

**Figure 11.** Relationship between the material constants and the W/B for 0% or 5% LCDBA replacement series, corresponding to Figure 9; (**a**) constant *a* (**b**) constant *b*.

#### **4. Conclusions**

This experimental study evaluates the LCD by-product of the refining process, LCDBA, as an alternative and sustainable alkaline activator for GGBFS blended cement concrete. To investigate the applicability of this the alternative activator, the tested experimental parameters were set at three LCDBA replacement levels and four water/binder ratios. The activation effects were characterized based on XRD and TG analyses using GGBFS blended cement paste. The fresh and hardened properties of GGBFS blended cement concrete incorporating LCDBA were investigated using slump, bleeding, setting time, and compressive strength tests. The key observations and findings of this research can be summarized as follows:


(4) The series of compressive strength tests conducted in this research concluded that LCDBA was an effective alkaline activator for GGBFS blended cement concrete, showing early-age strength developing characteristics, especially for mixtures with lower water/binder ratios. The compressive strength model, suggested by the ACI Committee 209, also highlighted the early-age strength developing characteristics of the blended concrete activated with LCDBA.

The results obtained in this study provide a simple, yet effective and practical, means of reusing an industrial by-product as an alternative alkaline activator for GGBFS blended concrete. Further studies are, however, necessary to determine its long-term durability and dimensional stabilities, such as shrinkage and creep.

**Author Contributions:** Conceptualization, S.C.; methodology, S.C.; software, S.C.; validation, S.C.; formal analysis, S.C. and S.P.; investigation, S.C.; resources, S.P.; data curation, S.C.; writing—original draft preparation, S.C.; writing—review and editing, S.P.; visualization, S.C.; supervision, S.P.; project administration, S.P.; funding acquisition, S.P. Both authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (no. NRF-2019R1F1A1060906). The research described herein was also supported by the 2019 Research Fund (1.190015.01) of UNIST (Ulsan National Institute of Science and Technology).

**Conflicts of Interest:** The authors declare no conflict of interest.
