Accelerated alkali-silica reaction after a seven-year ASR-dormancy period

Abstract

Keywords: Alkali silica reaction; durability; cement; supplementary cementitious materials

1. Introduction

Alkali silica reaction is one of concrete's significant and commonly seen durability problems; therefore, innumerable studies have been carried out for decades worldwide to understand its nature and overcome its deleterious effects. Since Stanton's paper [1], dealt with the chemistry of ASR, many test methods have been developed to analyze the reactivity of aggregates and to assess the effectiveness of various ASR-mitigation approaches. However, there are still many ASR-suffered structures and many at risk of being suffered. Therefore, the repair and maintenance works—techniques include, but are not limited to, injection of epoxy resins into cracks [2], precast prestressed concrete confinement [3], CFRP sheet bonding [4], surface coating using acrylic

and urethane-based continuous fiber sheets, and concrete jacketing [5] as well as novel methods such as applying bacteria-containing grout to the surface of ASR-suffered concrete [6]—have been carried out for many years worldwide and will be executed in the future. A successful repair must be accompanied either by a waterproofing application or drainage/removal of water in contact with the concrete. In other words, the dryness of the ASR-suffered concrete must be ensured— such as improvement of drainage, sealing surface of the structure to prevent water ingress, and silane treatment of the structure [7]—to halt reactions (ASR-dormancy period) and to avoid swelling of the existing ASR gel. Otherwise, the reaction and the expansion will continue; thus, the deterioration will occur again. Daidai et al. [3] have reported that the water trapped

*Corresponding author. *E-mail address: ahsan.beglari@okan.edu.tr / ahsan.beglari@gmail.com Published by Yıldız Technical University Press, İstanbul, Türkiye This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 327–334, December 2024

Table 1. The additive content of the Portland cement used in this study Cement type

inside the concrete before the injection repair and surface coating works was the main reason behind the continuous expansion of the repaired bridge concrete. They have also emphasized that penetration of the drainage water and deicing salts to the repaired concrete bridge piers accelerated the ASR reactions. The ASR-dormancy period can be maintained as long as the concrete is dry. The ASR-dormancy period established by preventing water ingress will be over whenever water penetrates any ASR-suffered concrete. To the best of the author's knowledge, only one scientific paper dealt systematically with the post-dormancy (after 14 14-month drying period) ASR expansions [8]. Multon and Toutlemonde [8] have investigated the effect of moisture conditions and transfers on ASR-suffered concrete specimens. They showed that water supply causes new ASR expansions in the ASR-suffered dry concretes. However, there is a lack of information regarding the effects of pre-dormancy expansion levels and supplementary cementitious material usage on post-dormancy expansions. This study endeavored to understand this matter through accelerated reactions and a seven-year ASR-dormancy period (7Y-DP). It can also shed light on using recycled concrete aggregates obtained from ASR-suffered concrete structures.

and 0.5% expansions (14-day) in accelerated mortar bar test. The mineralogical and microstructural characteristics of the aggregate were studied in detail by Copuroglu et al. [15]. They revealed that the primary source of the observed expansion can be explained by the reactive glassy phase of the basalt matrix having approximately 70% SiO2 [15].

3. Results And Discussion

In the ASTM C 1260 method, aggregate is usually considered reactive if the cement-aggregate combinations (mortar bars) immersed in the 80 oC NaOH solution exhibit expansions greater than 0.10% after 14 days of exposure

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 327–334, December 2024

Table 2. The chemical compositions of the cement and the SCMs CaO MgO Na2O K2O Cl SO3 Free SiO2 Al2O3 Fe2O3 CaO

Cemii-4

(CEM II/A-LL 42,5 R) CF (Class-C fly ash) FF (Class-F fly ash)

according to ASTM C 1778-20 [19]. In addition, combinations of SCMs and aggregates (ASTM C 1567) [17] that result in an expansion of less than 0.10% after 14 days are considered an "acceptable" level. As mentioned earlier, the ASR-mitigation ability of SCMs, after all now, is a well-known phenomenon; therefore, this study endeavors to focus on the further expansion of various specimens following a relatively long ASR-dormant period. However, the 14-day expansions should be briefly evaluated. Despite the concerns regarding the severe test conditions of the accelerated mortar bar (AMB) test, Bérubé et al. [20] have emphasized that it can be applied for evaluation of the ASR-mitigation performance of SCMs as long as the test period is restricted to 14 days. As shown in Figure 1, the performance of the blended cement, type II (CEMII), depends, obviously, on their additive contents. The CEMII-1 and CEMII-2 had significantly lower— of course, they were not below the critical level— expansions compared to those of the CEMs, while the expansions of the 3rd and 4th CEMIIs were almost the same with the CEMIs. Given the contents of the blended cement, these results are expected outcomes; the CEMII-1 and 2 include 22–24% fly ash, but none in the CEMII-4 and only ∼7% in the CEMII-3. The utilization of the FF and S had almost similar effects and could keep the expansion of the mortars, no matter the cement type, below the critical level. However, the CF couldn't do so. In other words, the 30% and 25% replacement ratios were insufficient for the class C fly ash. The

Figure 1. The expansions of the mortar bars before and after the 7Y-DP. relatively poorer performance of class C fly ashes is not an unexpected nor unknown outcome [11, 21].

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 327–334, December 2024

Figure 2. The expansions of the mortar bars before and after the 7Y-DP. The expansions of the mortar bars after the 7Y-DP will be evaluated through two different types of graphs: 1) Ignoring the expansion values before the 7Y-DP as if the specimen were subjected to the NaOH solution for the first time, 2) Continuous (cumulative expansion) curves; as if there was no a dormancy period. Figure 2 shows the expansions of the mortar bars before and after the 7Y-DP. It is obvious that the pre-7Y-DP (48 days) reactions considerably reduced the post-7Y-DP expansions of all CEMI mixtures as compared to the pre-7YDP's expansions — as pointed out earlier, this 7-year can be considered as a period in which ASR reactions is halted, for example, by preventing water ingress. This behavior is also valid for the plain CEMII mixtures; nevertheless, the extent of the decrement, as expected, is not high for the CEMII-1 and CEMII-2 mixtures—because the pre-7Y-DP expansions were lower than those of the other plain CEMI/CEMII mixtures. It must be mentioned that the CEMII-1 and 2 include 22–24% fly ash, but none in the CEMII-4 and only ∼7% in the CEMII-3. Generally speaking, the higher the pre-7Y-DP expansion of a mixture reaches, the lower post-7Y-DP expansion would be—of course, compared to its pre-7Y-DP expansions. This fact is much more apparent in the plain CEMI mixtures with no ASR-mitigation mechanism. In other words, a great extent of the possible accelerated ASR reactions/or expansions—given the severe test condition of the AMB test, it is not necessary to have the same expansion

level in the site even up to the end of service life of a real structure—that can be observed in these mixtures within a certain period had already occurred prior to the 7Y-DP. The source of the post-7Y-DP expansions is the swelling of the old gel (absorption of water by the dry gel) and further ASR reaction, thus forming a new gel. The swelling of the old gel is more likely to be the dominant mechanism in the mixtures with higher pre-7Y-DP expansions, or at least, the contribution ratio of the post-7Y-DP expansions was higher in such mixtures. In other words, the residual available reactive silica would be limited in the mixtures in which there was no or insufficient SCM. Multon and Toutlemonde [8] showed that absorption of water (re-wetting) by old gel at any time of the life of an ASR-damaged structure can cause rapid swelling. This phenomenon was also reported in a few studies that dealt with the ASR reactivity of recycled concrete aggregates (RCA) obtained from ASR-suffered concrete. The expansion of old gel in the RCA is considered one of the contributing factors to the higher expansion (concrete prism test) of the RCA-containing concrete compared to the concrete containing the same virgin reactive aggregate [22, 23]. In addition, one of the reasons behind the lower expansion of the same RCA in AMBT was attributed to the loss of the old gel—that would swell by absorbing the moisture—as a result of further crushing of the coarse RCA to produce sand-size fraction for the test [22, 24].

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 327–334, December 2024

Figure 3. The cumulative expansions of the mortar bars before and after the 7Y-DP. The post-7Y-DP expansions of S-containing and FF-containing mixtures were almost the same level as those of the pre-7Y-DP (43-day), even though the pre-7Y-DP exposure period (57–59 days) of them was 9–11 days longer than that of plain (48 days) mixtures. This finding shows that a significant amount of the ASR reaction/expansion was delayed owing to the ASR-mitigation effect of the S and FF; therefore, the specimens expanded almost like the pre7Y-DP —except the first 14 days in which the ASR-mitigation effect of SCMs are much more dominant (it will be discussed later)— as if they were subjected to the NaOH solution for the first time and did not have an ASR history. It can be assumed that the contribution ratio of the new ASRgel (post-7Y-DP gel) on the expansion was higher in these mixtures. On the other hand, the CF-containing mixtures acted almost like the plain mixtures because of the poorer ASR-mitigation performance of the class C fly ash. In other words, the 30% and 25% CF replacement ratios were not sufficient to delay the ASR reaction of, respectively, the CEMI and CEMII mixtures; therefore, a great extent of the possible accelerated ASR reactions/or expansions that can be observed in these mixtures within a certain period of time had already occurred before the 7Y-DP. It should be noted that the pre-7Y-DP exposure period (42 days) was even one day shorter than that of the post-7Y-DP; it was 15 and 17 days shorter than the post-7Y-DP exposure periods of S-containing and FF-containing mixtures, respectively.

The abovementioned behavior would be more evident if the pre-7Y-DP exposure period were extended. An interesting point comes out when the post-7Y-DP expansions are cumulatively added to the pre-7Y-DP curves as if there was no dormancy period (Fig. 3). Meanwhile, it should be noted that ASR-induced expansion is permanent and cumulative in the presence of water [8]. Multon and Toutlemonde [8] have reported that the shrinkage of ASR-suffered concrete due to drying at 30% RH had perfectly the same kinetics and range— no matter the conditions of ASR-induced damage and the direction of measurements— as for the sound concrete. They have stated that the drying shrinkage cannot balance the expansion; thus, the expansions and cracks are irreversible [8]. As shown in Figure 3, the appearance of the cumulative curves is almost like continuous ASR expansions, as if there were no 7Y-DP—especially S and FF-containing mixtures. Moreover, the increasing rates of the initial portion of the post-7Y-DP expansions almost match the pre-7Y-DP expansion trends. This behavior can be considered solid proof for the finding of Multon and Toutlemonde [8], who showed that water absorption by the dry old gel at any time can cause rapid swelling. If not, the formation of new gel— without the expansion of the old gel— would cause lower expansion rates. This phenomenon is much more critical for the mixtures in which there was no or insufficient SCM. As pointed out earlier, the residual available reactive silica

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 327–334, December 2024

Cemii-4

would be limited in these mixtures for the post-7Y-DP period, so the source of such post-7Y-DP expansion rates could not, certainly, be just the formation of a new gel. As mentioned earlier, a significant amount of the ASR reaction was delayed owing to the ASR-mitigation effect of the S and FF; therefore, it can be assumed that the contribution ratio of the new ASR-gel (post-7Y-DP gel) on the expansion was higher in these mixtures. The reason behind the same increasing rate of the very initial portion of the post-7Y-DP expansions as the very last portion of the pre-7Y-DP expansions can be explained not only by the rapid swelling of the old gel but also by the early formation of new ASR gel— that no longer could be suppressed by the ASR-mitigation effect of SCMs in AMBT (80 oC NaOH solution). It should be remembered that AMBT can be applied to evaluate the ASR-mitigation performance of SCMs as long as the test period is restricted to 14 days [20]. There are vital differences between the ASR mitigation mechanisms of SCMs in the AMBT and those in other test methods. Reduction in the permeability of the matrix due to the pozzolanic reaction of SCMs is the dominant ASR-mitigation mechanism in the AMBT [25] as the solution continually supplies alkali. Therefore, higher impermeability strengthens the cementitious matrix's resistance against the ions' diffusion [25]. Diminution in the alkalinity of the matrix itself is, of course, one of the mechanisms, but it is not the main reason, as in the concrete prism methods [26]. Such a high impermeability was no longer the case beyond 14 days due to the continuous formation of cracks in the very early post-7Y-DP ASR reactions. Thus, the formation of new gel would be almost like that in the mixtures with insufficient SCMs. Although in natural structures and realistic conditions, such high expansion is not likely to occur in the SCM-containing mixtures, these results show that the ASR-mitigation capacity of SCMs would be limited after the repairing work conducted

on a structure deteriorates under extreme conditions. But, of course, the primary purpose of using SCMs in this study was to scrutinize the effects of the pre-dormancy expansion levels on the post-dormancy expansions. Beyond 14-day, as mentioned earlier, AMBT is not a suitable test method for analyzing the ASR-mitigation behavior of SCMs; in future studies, the effect of dormancy period on the ASR-mitigation behavior could be investigated in the test methods having more realistic test conditions like concrete prism test. In addition to assessing the initial increasing rates and the appearance of the curves, a series of statistical analyses were also conducted on each of the 32 mixtures to understand better the overall trends of the pre-7Y-DP and post-7YDP expansion curves. The Levene test was applied to measure the homogeneity of variances between the pre-7Y-DP and post-7Y-DP period data. According to the Levene test results, since "p" was less than 0.05 in all mixtures (Table 3), it was proven that the variances of both periods were homogeneous at the 95% significance level. Then, an analysis of covariance (ANCOVA) was applied between both periods. In the ANCOVA test, the significance of "Period—Time" (P-T) interaction was used to assess whether the effect of time on a dependent variable, ASR expansions, differs across the periods. This interaction term captures any systematic change in the trend or rate between the two periods. If P (significance) is less than 0.05, it suggests that the relationship between time and expansion differed between periods. A significant P-T interaction would indicate that ASR trends changed after the 7Y-DP. In this case, either the dormancy or the pre-7Y-DP expansion/reaction levels —that is more likely to be, as discussed earlier in detail— might have altered the trend of post-7Y-DP expansions relative to that of pre-7Y-DP. On the other hand, no significant P-T interaction (p≥0.05) suggests that the ASR cumulative expansion curve continued after the 7Y-DP without substantial alter-

J Sustain Const Mater Technol, Vol. 9, Issue. 4, pp. 327–334, December 2024

ation. In other words, in such mixtures, the ASR reaction was delayed to a great extent —owing to the ASR-mitigation effect of some of the SCMs— so that it could be interpreted that the contribution ratio of the new ASR-gel (post-7Y-DP gel) on the expansion was much higher in these mixtures. The P-T interaction's Sig. (p) values obtained from ANCOVA analysis of each mixture's data are presented in Table 3. The SCM-containing mixtures highlighted by the black color in the table have no significant P-T interaction (p>0.05), implying no statistically meaningful difference in slopes between the periods. In other words, the ASR reaction was delayed to a great extent, so the contribution ratio of the new ASR-gel (post-7Y-DP gel) to the expansion was much higher in these mixtures than in the other SCM-containing mixtures. As can be seen, this phenomenon was dominant only in the GGBFS-containing mixtures.

References

1. 1. Stanton, T. E. (1942). Expansion of concrete through reaction between cement and aggregate. Trans Am Soc Civ Eng, 107(1), 54-84. [CrossRef]
2. 2. Daidai, T., Torii, K. (2008). A proposal for rehabilitation of ASR-affected bridge piers with fractured steel bars. Proc. of 13th Inter. Conf. on Alkali-aggregate Reaction in Concrete.
3. 3. Daidai, T., Andrade, O., Torii, K. (2012). The maintenance and rehabilitation techniques for ASR-affected bridge piers with fracture of steel bars. Proc. 14th Inter. Conf. on Alkali-aggregate Reaction in Concrete, Austin, Texas.
4. 4. Torii, K., Prasetia, I., Minato, T., Ishii, K. (2012). The feature of cracking in prestressed concrete bridge girders deteriorated by alkali-silica reaction. Proc. 14th Inter. Conf. on Alkali-aggregate Reaction in Concrete, Austin, Texas.
5. 5. Nomura, M., Komastubara, A., Fujimoto, K., Torii, K. (2013). Evaluation of maintenance methods for ASR-damaged structures in Hokuriku district, Japan. Proceedings, Third International Conference on Sustainable Construction Materials and Technologies, Kyoto, Japan.
6. 6. Luan, Y., Arasawa, T., Mutsuyoshi, H., & Kawana, R. (2021). Repair of alkali-silica reaction-induced cracks using bacteria: Crack recovery and other properties. ACI Mater J, 118(4), 133-142. [CrossRef]
7. 7. Bakker, J. D. (2008). Control of ASR related risks in the Netherlands. Proceedings of the 13th International Conference on Alkali-Aggregate Reaction in Concrete, Trondheim, Norway, (pp. 16-20).
8. 8. Multon, S., & Toutlemonde, F. (2010). Effect of moisture conditions and transfers on alkali silica reaction damaged structures. Cement Concr Res, 40(6), 924-934. [CrossRef]
9. 9. ASTM International. (2019). Standard specification for coal fly ash and raw or calcined natural pozzolan for use in concrete. ASTM-C618.
10. 10. Yüksel, C., Mardani-Aghabaglou, A., Beglarigale, A., Yazıcı, H., Ramyar, K., & Andiç-Çakır, Ö. (2017). Alkali-silica reaction expansions and the extent of alkali leaching in concretes containing basalt and waste glass as aggregate. Teknik Derg, 28(2), 7865-7882. [CrossRef]
11. 11. Yüksel, C., Mardani-Aghabaglou, A., Beglarigale, A., Yazıcı, H., Ramyar, K., & Andiç-Çakır, Ö. (2016). Influence of water/powder ratio and powder type on alkali-silica reactivity and transport properties of self-consolidating concrete. Mater Struct, 49, 289-299. [CrossRef]
12. 12. Andiç-Çakır, Ö. (2007). Investigation of test methods on alkali aggregate reaction, [Doctoral dissertation], Ege University.
13. 13. Beglarigale, A., Yazıcı, H. (2013). The effect of alkali-silica reaction on steel fiber-matrix bond characteristics of cement based mortars. Constr Build Mater, 47, 845-860. [CrossRef]
14. 14. Beglarigale, A., & Yazici, H. (2014). Mitigation of detrimental effects of alkali-silica reaction in cement-based composites by combination of steel microfibers and ground-granulated blast-furnace slag. J Mater Civ Engin, 26(12), 04014091. [CrossRef]
15. 15. Çopuroğlu, O., Andiç-Çakir, Ö., Broekmans, M. A., & Kühnel, R. (2009). Mineralogy, geochemistry and expansion testing of an alkali-reactive basalt from western Anatolia, Turkey. Mater Charact, 60(7), 756-766. [CrossRef]
16. 16. ASTM International. (2007). 1260 Standard test method for potential alkali reactivity of aggregates (Mortar-Bar Method). ASTM C1260-07.
17. 17. ASTM International. (2007). Standard test method for determining the potential alkali-silica reactivity of combinations of cementitious materials and aggregate (Accelerated Mortar-Bar Method). ASTM C1567-07.
18. 18. Turkish Standards Institution. (2012). Complementary Turkish Standard for the implementation of TS EN 206. TS 13515.
19. 19. ASTM International. (2020). Standard guide for reducing the risk of deleterious alkali-aggregate reaction in concrete. ASTM C1778-20.
20. 20. Bérubé, M. A., Duchesne, J., & Chouinard, D. (1995). Why the accelerated mortar bar method ASTM C1260 is reliable for evaluating the effectiveness of supplementary cementing materials in suppressing expansion due to alkali-silica reactivity. Cement Concrete Aggregates, 17(1), 26-34. [CrossRef]
21. 21. Shon, C. S., Sarkar, S. L., & Zollinger, D. G. (2004). Testing the effectiveness of Class C and Class F fly ash in controlling expansion due to alkali-silica reaction using modified ASTM C1260 test method. J Mater Civ Eng, 16(1), 20-27. [CrossRef]
22. 22. Johnson, R., & Shehata, M. H. (2016). The efficacy of accelerated test methods to evaluate alkali silica reactivity of recycled concrete aggregates. Constr Build Mater, 112, 518-528. [CrossRef]
23. 23. Piersanti, M., & Shehata, M. H. (2022). A study into the alkali-silica reactivity of recycled concrete aggregates and the role of the extent of damage in the source structures: Evaluation, accelerated testing, and preventive measures. Cem Concr Compos, 129, 104512. [CrossRef]
24. 24. Piersanti, M. (2015). Testing recycled concrete aggregate suffering different levels of alkali-silica reaction for use in new structures [Doctoral dissertation], Toronto Metropolitan University.
25. 25. Xu, G. J., Watt, D. F., & Hudec, P. P. (1995). Effectiveness of mineral admixtures in reducing ASR expansion. Cem Concr Res, 25(6), 1225-1236. [CrossRef]
26. 26. Kawabata, Y., & Yamada, K. (2015). Evaluation of alkalinity of pore solution based on the phase composition of cement hydrates with supplementary cementitious materials and its relation to suppressing ASR expansion. J Adv Concr Technol, 13(11), 538-553. [CrossRef]

Share and Cite