Inhibition of Alkali-Carbonate Reaction by Fly Ash and Metakaolin on Dolomitic Limestones

In this paper, the dolomitic limestone determined as alkali–carbonate-reactive by various methods is used as an aggregate. Inhibition experiments were carried out on the basis of the concrete microbar method (RILEM AAR-5 standard), in which 10%, 30%, and 50% fly ash and metakaolin were used to replace cement. Thermogravimetric–differential scanning calorimetry (TG-DSC), X-ray diffractometry (XRD), mercury intrusion porosimetry (MIP), and scanning electron microscopy–energy dispersive X-ray spectrometry (SEM-EDS) were used to analyze the inhibition mechanism of fly ash and metakaolin on ACR. The results show that the expansion of samples at the age of 28 days are less than 0.10% when the fly ash contents exceed 30% and the metakaolin contents exceed 10%, which proves that the ACR is inhibited effectively. Meanwhile, the Ca(OH)2 content of the samples was reduced and the pore structure of the samples was optimized after adding fly ash and metakaolin. The dolomite crystals in the samples containing 50% fly ash and metakaolin are relatively complete.


Introduction
The alkali-aggregate reaction (AAR) includes the alkali-silica reaction (ASR) and the alkali-carbonate reaction (ACR), ACR is a reaction that occurs between the carbonate components of the aggregate and the alkaline solution, resulting in anomalous expansion and disintegration of the concrete. In 1957, Swenson [1,2] first discovered the case of engineering damage caused by ACR. At that time, he found that some concretes using carbonate aggregates in Kingston, Canada produced severe network cracks. As the microscopic characteristics of the cracks produced by these concretes were very different from those caused by ASR, he proposed that the carbonate component contained in the aggregate would react with the alkaline solution and cause cracking of the concrete structure. As a result, scholars from various countries have studied ACR, and a series of ACR expansion mechanisms have been proposed.
Since the ACR was proposed, there has been a relatively consistent view [3]. The ACR is a chemical reaction between the alkali (K + , Na + ) in the pore solution and the dolomite (CaMg(CO 3 ) 2 ) in the aggregate (also known as a dedolomization reaction), and the reaction equation is as follows: CaMg(CO 3 ) 2 + 2MOH → Mg(OH) 2 + CaCO 3 + M 2 CO 3 (1) where M represents alkali metal ions: K + , Na + , or Li + M 2 CO 3 + Ca(OH) 2 → 2MOH + CaCO 3 , Reaction Formula (1) is the dedolomitization reaction; Formula (2) indicates that M 2 CO 3 continues to react with Ca(OH) 2 to form calcite and regenerate MOH. Metakaolin (MK) is produced by Super Kaolin Co. Ltd. Inner Mongolia, China. It is calcined and ground from kaolin ore. Table 1 shows its chemical composition. MK is mainly in the form of amorphous aluminum silicate. The morphology of metakaolin observed by scanning electron microscope (SEM) is shown in Figure 1.
Class F fly ash (FA) from Nanjing Pudi, Table 1 shows its chemical composition. The particle morphology of fly ash observed by scanning electron microscope (SEM) is shown in Figure 1. porosimetry (MIP), and scanning electron microscopy-energy dispersive X-ray spectrometry (SEM-EDS) were used to analyze the inhibition mechanism of fly ash and metakaolin on ACR. The study will guide the practical application of dolomite in concrete structures.

Materials
2.1.1. Cementitious Materials P·II 52. 5 Portland cement (OPC) from Jiangnan-Onada Cement Co. Ltd. Nanjing, China was used in this experiment. Its alkali content was 0.56%. The chemical composition of the cement was analyzed by X-ray fluorescence (XRF) in Table 1.  Table 1 shows its chemical composition. MK is mainly in the form of amorphous aluminum silicate. The morphology of metakaolin observed by scanning electron microscope (SEM) is shown in Figure 1.
Class F fly ash (FA) from Nanjing Pudi, Table 1 shows its chemical composition. The particle morphology of fly ash observed by scanning electron microscope (SEM) is shown in Figure 1.

Aggregates
The aggregate is Ordovician dolomitic limestone S1# derived from Baofuling, Shandong Province, China. Table 2 shows its chemical composition. The XRD patterns of dolomitic limestone S1# are demonstrated in Figure 2. The dolomitic limestone S1# is mainly composed of calcite, dolomite, and quartz. The typical microstructure of dolomitic limestones by is shown in Figure 3; dolomite crystals in dolomitic limestone S1# are dispersed in the calcite matrix.

Aggregates
The aggregate is Ordovician dolomitic limestone S1# derived from Baofuling, Shandong Province, China. Table 2 shows its chemical composition. The XRD patterns of dolomitic limestone S1# are demonstrated in Figure 2. The dolomitic limestone S1# is mainly composed of calcite, dolomite, and quartz. The typical microstructure of dolomitic limestones by is shown in Figure 3; dolomite crystals in dolomitic limestone S1# are dispersed in the calcite matrix.

Concrete Microbars Test
According to the standard of RIELM AAR-5, the experiment used dolomitic limestone S1# as aggregate (around 5 mm-10 mm) and Portland cement (alkaline content adjusted to 1.25% with NaOH) to prepare concrete microbars (40 mm × 40 mm × 160 mm), the cement to aggregate ratio is 1:1, and the water-cement ratio is 0.32. Fly ash and metakaolin replaced 10%, 30%, and 50% of cement quality respectively. Table 3 shows a mixed design of concrete microbars. The concrete microbars were cured in a solution of 1M NaOH at 80 • C, and the expansion of the concrete microbars were measured regularly. If the expansion of the concrete microbars at age of 28 days is less than 0.1%, it is considered that ACR can be inhibited.

Determination of Hydration Products
Thermogravimetric-differential scanning calorimetry (TG-DSC) and X-ray diffractometry (XRD) were used to determine the change of Ca(OH) 2 content in the concrete.
Cement pastes containing 10%, 30%, and 50% fly ash and metakaolin were made into about 2-3 mm cubes. The TG-DSC test used the STA 449 F3 thermal analyzer from NETCSZH Company, Bayern, Germany. Samples were heated from 0 to 600 • C at a heating rate of 10 • C/min in a N 2 atmosphere, and α-Al 2 O 3 was used as the reference to analyze the heat and mass changes of the samples during the heating process.
The cement pastes containing 10%, 30%, and 50% fly ash and metakaolin were dried, ground with an agate mortar, the particle size was controlled below 80 µm, and the samples filled into a 20 × 20 × 0.5 mm quartz sample tank. The XRD test used the SmartLab (3) X-ray diffractometry (Cu target, rated power 3 kW, scanning range of 5 • -80 • with a step size 0.02 • and scanning speed of 10 • /min) from Rigaku Company, Tokyo, Japan.

Mercury Porosimetry Analyses
A cube about 2-3 mm in the center of the concrete microbars containing 10%, 30%, and 50% fly ash and metakaolin was cut and PoreMaster-60 mercury porosimeter was used to analyze the pore structure of the selected samples. The test has two phases-low pressure from 0 psi to 25 psi (0.172 MPa) and high pressure from 20 psi (0.138 MPa) to 50,000 psi (344.738 MPa).

Microscopic Analysis of ACR Products
Ultra55 field emission scanning electron microscopy from Zeiss Company Jena, Germany was used to observe the morphology of the ACR products in concrete microbars. Table 4 shows the expansion of accelerated mortar bars (ASTM C1260), concrete microbars (RILEM AAR-5), concrete prisms (ASTM C1293), and rock prisms (ASTM C586). The accelerated mortar bars method was used to determine the alkali-silica reactivity of dolomitic limestones-if the expansion of the mortar bars at the age of 14 days exceeds 0.1%, the aggregates are classified as potentially alkali-silica-reactive. The concrete microbars method, concrete prisms method, and rock prisms were used to determine the alkalicarbonate reactivity of dolomitic limestones. If the expansion of the concrete microbars at age of age of 28 days exceeds 0.10%, the aggregates are classified as alkali-carbonatereactive. If the expansion of the concrete prisms at the age of 1 year exceeds 0.04%, the aggregates are classified as alkali-carbonate-reactive. If the expansion of the rock prisms at age of 84 days exceeds 0.10%, the aggregates are classified as alkali-carbonate-reactive. Four methods were used to prove that aggregate S1# were classified as alkali-carbonate-reactive and not alkali-silica-reactive.

Inhibition of ACR with Fly Ash
In order to study the effect of fly ash on inhibiting ACR, 10%, 30%, and 50% fly ash was used to replace cement. The expansion rates of concrete microbars at the age of 28 days are shown in Figure 4.
at age of age of 28 days exceeds 0.10%, the aggregates are classified as alkali-carbonatereactive. If the expansion of the concrete prisms at the age of 1 year exceeds 0.04%, the aggregates are classified as alkali-carbonate-reactive. If the expansion of the rock prisms at age of 84 days exceeds 0.10%, the aggregates are classified as alkali-carbonate-reactive. Four methods were used to prove that aggregate S1# were classified as alkali-carbonatereactive and not alkali-silica-reactive.

Inhibition of ACR with Fly Ash
In order to study the effect of fly ash on inhibiting ACR, 10%, 30%, and 50% fly ash was used to replace cement. The expansion rates of concrete microbars at the age of 28 days are shown in Figure 4. Age/d S1# FA0% S1# FA10% S1# FA30% S1# FA50%  Figure 4 shows that the expansion of the control samples continued to develop with age, and the expansion at age of 28 days was 0.177%. The fly ash reduced the ACR expansion of the samples, and the higher the fly ash content, the smaller the expansion. The expansion of the concrete microbars containing 10% fly ash at age of 28 days was 0.129% which is still higher than the defined threshold (0.1%), which indicates that the low proportion of fly ash has poor ACR inhibition effect on the samples. When the content of fly ash increased to 30% and 50%, the expansion of concrete microbars at age of 28 days decreased to 0.051% and 0.014%. The ACR expansion of the samples was significantly inhibited. The results show that only the fly ash with high mass fractions can significantly inhibit the ACR of the samples, while the fly ash with low mass fraction cannot inhibit the ACR of the samples effectively.  Figure 4 shows that the expansion of the control samples continued to develop with age, and the expansion at age of 28 days was 0.177%. The fly ash reduced the ACR expansion of the samples, and the higher the fly ash content, the smaller the expansion. The expansion of the concrete microbars containing 10% fly ash at age of 28 days was 0.129% which is still higher than the defined threshold (0.1%), which indicates that the low proportion of fly ash has poor ACR inhibition effect on the samples. When the content of fly ash increased to 30% and 50%, the expansion of concrete microbars at age of 28 days decreased to 0.051% and 0.014%. The ACR expansion of the samples was significantly inhibited. The results show that only the fly ash with high mass fractions can significantly inhibit the ACR of the samples, while the fly ash with low mass fraction cannot inhibit the ACR of the samples effectively.

Inhibition of ACR with Metakaolin
In order to study the effect of metakaolin on inhibiting ACR, 10%, 30%, and 50% metakaolin was used to replace cement. The expansion rates of concrete microbars at the age of 28 days are shown in Figure 5. Figure 5 shows that metakaolin has an excellent inhibitory effect on ACR. The expansion of concrete microbars at age of 28 days with metakaolin contents of 10%, 30%, and 50% were 0.081%, 0.020%, and 0.007%, which were all lower than the defined threshold (0.1%). Compared with the control samples, 10% metakaolin can effectively inhibit the ACR of the samples. The inhibitory effect of 30% and 50% metakaolin is more obvious. The results show that metakaolin is highly effective in inhibiting ACR.

Inhibition of ACR with Metakaolin
In order to study the effect of metakaolin on inhibiting ACR, 10%, 30%, and 50% metakaolin was used to replace cement. The expansion rates of concrete microbars at the age of 28 days are shown in Figure 5. Expansion / % Age/d S1# MK0% S1# MK10% S1# MK30% S1# MK50% Figure 5. The expansion of the concrete microbars with metakaolin. Figure 5 shows that metakaolin has an excellent inhibitory effect on ACR. The expansion of concrete microbars at age of 28 days with metakaolin contents of 10%, 30%, and 50% were 0.081%, 0.020%, and 0.007%, which were all lower than the defined threshold (0.1%). Compared with the control samples, 10% metakaolin can effectively inhibit the ACR of the samples. The inhibitory effect of 30% and 50% metakaolin is more obvious. The results show that metakaolin is highly effective in inhibiting ACR.

Effects of Fly Ash and Metakaolin on Ca(OH)2 in Cement Paste
As one of the important hydration products of cement, Ca(OH)2 has been recognized by many scholars for its role in maintaining the alkalinity of the cement system and promoting AAR.

XRD Analyses
X-ray diffraction analysis can semi-quantify the change of Ca(OH)2 content in cement paste. Figure 6 are XRD patterns of cement pastes containing 0%, 10%, 30%, and 50% fly ash and metakaolin that were cured at 20 °C for 28 days. With the increase of the content of fly ash and metakaolin, the diffraction peak intensity of Ca(OH)2 in the XRD pattern also decreases, which indicates that the content of Ca(OH)2 in the cement paste decreases with the increase of the content of fly ash and metakaolin.

Effects of Fly Ash and Metakaolin on Ca(OH) 2 in Cement Paste
As one of the important hydration products of cement, Ca(OH) 2 has been recognized by many scholars for its role in maintaining the alkalinity of the cement system and promoting AAR.

XRD Analyses
X-ray diffraction analysis can semi-quantify the change of Ca(OH) 2 content in cement paste. Figure 6 are XRD patterns of cement pastes containing 0%, 10%, 30%, and 50% fly ash and metakaolin that were cured at 20 • C for 28 days. With the increase of the content of fly ash and metakaolin, the diffraction peak intensity of Ca(OH) 2 in the XRD pattern also decreases, which indicates that the content of Ca(OH) 2 in the cement paste decreases with the increase of the content of fly ash and metakaolin.  Figure 6. XRD patterns of cement pastes containing 0%, 10%, 30%, and 50% fly ash and metakaolin that were cured at 20 °C for 28 days.

TG-DSC Analyses
TG-DSC can accurately determine the content of Ca(OH)2 in cement paste, which can be accurate to less than 0.1%. Figure 7 is TG-DSC curves of cement pastes containing 0%, 10%, 30%, and 50% fly ash that were cured in an aqueous solution at 20 °C for 28 days. It can be clearly found from the curves that the intensity of the Ca(OH)2 endothermic peak in the cement pastes decreases significantly with the increase of the fly ash content, which indicates that the content of Ca(OH)2 in the cement paste decreases with the increase of fly ash content. According to the TG-DSC curve, the content of Ca(OH)2 in cement paste can be calculated. The contents of Ca(OH)2 in cement paste containing 0%, 10%, 30%, and 50% fly ash were 12.3%, 11.5%, 9.0%, and 7.4% respectively. Figure 6. XRD patterns of cement pastes containing 0%, 10%, 30%, and 50% fly ash and metakaolin that were cured at 20 • C for 28 days.

TG-DSC Analyses
TG-DSC can accurately determine the content of Ca(OH) 2 in cement paste, which can be accurate to less than 0.1%. Figure 7 is TG-DSC curves of cement pastes containing 0%, 10%, 30%, and 50% fly ash that were cured in an aqueous solution at 20 • C for 28 days. It can be clearly found from the curves that the intensity of the Ca(OH) 2 endothermic peak in the cement pastes decreases significantly with the increase of the fly ash content, which indicates that the content of Ca(OH) 2 in the cement paste decreases with the increase of fly ash content. According to the TG-DSC curve, the content of Ca(OH) 2 in cement paste can be calculated. The contents of Ca(OH) 2 in cement paste containing 0%, 10%, 30%, and 50% fly ash were 12.3%, 11.5%, 9.0%, and 7.4% respectively.

TG-DSC Analyses
TG-DSC can accurately determine the content of Ca(OH)2 in cement paste, which can be accurate to less than 0.1%. Figure 7 is TG-DSC curves of cement pastes containing 0%, 10%, 30%, and 50% fly ash that were cured in an aqueous solution at 20 °C for 28 days. It can be clearly found from the curves that the intensity of the Ca(OH)2 endothermic peak in the cement pastes decreases significantly with the increase of the fly ash content, which indicates that the content of Ca(OH)2 in the cement paste decreases with the increase of fly ash content. According to the TG-DSC curve, the content of Ca(OH)2 in cement paste can be calculated. The contents of Ca(OH)2 in cement paste containing 0%, 10%, 30%, and 50% fly ash were 12.3%, 11.5%, 9.0%, and 7.4% respectively.  Figure 8 is TG-DSC curves of cement pastes containing 0%, 10%, 30%, and 50% metakaolin that were cured in an aqueous solution at 20 °C for 28 days. It can be clearly seen from the curves that the intensity of the Ca(OH)2 endothermic peak in the cement pastes decreases significantly with the increase of the fly ash content, which indicates that the content of Ca(OH)2 in the cement paste decreases with the increase of fly ash content. According to the TG-DSC curve, the content of Ca(OH)2 in the sample can be calculated. The contents of Ca(OH)2 in cement paste containing 0%, 10%, 30%, and 50% fly ash were 12.3%, 8.2%, 0%, and 0% respectively, which indicates that the Ca(OH)2 in cement pastes containing 30% and 50% metakaolin has been completely reacted.  Figure 8 is TG-DSC curves of cement pastes containing 0%, 10%, 30%, and 50% metakaolin that were cured in an aqueous solution at 20 • C for 28 days. It can be clearly seen from the curves that the intensity of the Ca(OH) 2 endothermic peak in the cement pastes decreases significantly with the increase of the fly ash content, which indicates that the content of Ca(OH) 2 in the cement paste decreases with the increase of fly ash content. According to the TG-DSC curve, the content of Ca(OH) 2 in the sample can be calculated. The contents of Ca(OH) 2 in cement paste containing 0%, 10%, 30%, and 50% fly ash were 12.3%, 8.2%, 0%, and 0% respectively, which indicates that the Ca(OH) 2 in cement pastes containing 30% and 50% metakaolin has been completely reacted. There are two main reasons for this result. On the one hand, the content of cement decreased after the fly ash and metakaolin partially replaced the cement, and the content of Ca(OH)2 generated by cement hydration also decreased. On the other hand, the secondary pozzolanic reaction of Ca(OH)2 with fly ash and metakaolin resulted in the decrease of Ca(OH)2 content in the system. With the hydration of cement, the fly ash and metakaolin react with hydration products Ca(OH)2 and C-S-H gel for secondary pozzolanic reaction, further producing C-S-H gel with low Ca/Si. The reduction of the content of Ca(OH)2 weakens the regeneration of alkali metal ions (Na + , K + ) and the secondary pozzolanic reaction produces a large amount of C-S-H gel with low Ca/Si, which has a stronger adsorption effect on alkali metal ions (Na + , K + ).

Effects of Fly Ash and Metakaolin on the Pore Structure of Concrete Microbars
The pore structure of concrete microbars cured in 80 °C , 1M NaOH solution for 28 days was analyzed. Figure 9 shows the pore size distribution and cumulative porosity of There are two main reasons for this result. On the one hand, the content of cement decreased after the fly ash and metakaolin partially replaced the cement, and the content of Ca(OH) 2 generated by cement hydration also decreased. On the other hand, the secondary pozzolanic reaction of Ca(OH) 2 with fly ash and metakaolin resulted in the decrease of Ca(OH) 2 content in the system. With the hydration of cement, the fly ash and metakaolin react with hydration products Ca(OH) 2 and C-S-H gel for secondary pozzolanic reaction, further producing C-S-H gel with low Ca/Si. The reduction of the content of Ca(OH) 2 weakens the regeneration of alkali metal ions (Na + , K + ) and the secondary pozzolanic reaction produces a large amount of C-S-H gel with low Ca/Si, which has a stronger adsorption effect on alkali metal ions (Na + , K + ).

Effects of Fly Ash and Metakaolin on the Pore Structure of Concrete Microbars
The pore structure of concrete microbars cured in 80 • C, 1M NaOH solution for 28 days was analyzed. Figure 9 shows the pore size distribution and cumulative porosity of concrete microbars containing 50% fly ash and metakaolin. The porosity of the blank sample was 21.64%, and the porosity of the sample containing 50% fly ash reached 22.41%. The porosity of the samples containing fly ash increased; however, the pore size distribution tends to be smaller. The porosity of the samples containing 50% metakaolin is 18.60%, and the porosity of the samples is reduced and the pore structure is more compact. The physical filling effect of fly ash and metakaolin and the filling effect of secondary pozzolanic reactants reduce the porosity of the cement stone, reduce the pore size, and densify the structure, which may be beneficial to prevent the diffusion of K + and Na + to the active aggregate and inhibit ACR.
of Ca(OH)2 content in the system. With the hydration of cement, the fly ash and metakaolin react with hydration products Ca(OH)2 and C-S-H gel for secondary pozzolanic reaction, further producing C-S-H gel with low Ca/Si. The reduction of the content of Ca(OH)2 weakens the regeneration of alkali metal ions (Na + , K + ) and the secondary pozzolanic reaction produces a large amount of C-S-H gel with low Ca/Si, which has a stronger adsorption effect on alkali metal ions (Na + , K + ).

Effects of Fly Ash and Metakaolin on the Pore Structure of Concrete Microbars
The pore structure of concrete microbars cured in 80 °C, 1M NaOH solution for 28 days was analyzed. Figure 9 shows the pore size distribution and cumulative porosity of concrete microbars containing 50% fly ash and metakaolin. The porosity of the blank sample was 21.64%, and the porosity of the sample containing 50% fly ash reached 22.41%. The porosity of the samples containing fly ash increased; however, the pore size distribution tends to be smaller. The porosity of the samples containing 50% metakaolin is 18.60%, and the porosity of the samples is reduced and the pore structure is more compact. The physical filling effect of fly ash and metakaolin and the filling effect of secondary pozzolanic reactants reduce the porosity of the cement stone, reduce the pore size, and densify the structure, which may be beneficial to prevent the diffusion of K + and Na + to the active aggregate and inhibit ACR.  Figure 10 is a SEM-EDS diagram of the inner products of the concrete microbars prepared by rock S1# after curing for 28 days at 80 • C in 1M NaOH. As can be seen, a large number of columnar (point 1) and lamellar (point 2) products are formed. EDS results show that the columnar products are mainly composed of Ca, C, and O, which is calcite. The lamellar products are mainly composed of Mg and O, which is brucite-this indicates that the dedolomization reaction also took place in the rock prisms. The length and width of the brucite are about 1 µm and the calcite particles are about 0.5 µm. The brucite and calcite stack each other, and there are a lot of pores between the products. The crystal size of brucite and calcite is larger, and the flaky brucite and calcite stacked together are not dense, and there are obvious pores. It can be seen that the reaction product gradually grows up and more and more brucite and calcite gather together in a limited space with the progress of the reaction. At the same time, the accumulation of products leads to the increase of pores between products, which causes the continuous expansion of the rock, which is also the reason for the expansion of concrete microbars. The results show that the expansion of concrete microbars cured in NaOH solution is caused by a dedolomization reaction. Figure 11 is a SEM-EDS diagram of the inner products of the concrete microbars containing 50% fly ash and metakaolin prepared by rock S1# after curing for 28 days at 80 • C in 1M NaOH. It can be seen from the figure that the dolomite crystals inside the rock are relatively complete, and no lamellar brucite and columnar calcite generated by the dedolomization reaction were found, which indicates that the addition of 50% fly ash and 50% metakaolin inhibit the progress of the dedolomization reaction. Figure 11 is a SEM-EDS diagram of the inner products of the concrete microbars containing 50% fly ash and metakaolin prepared by rock S1# after curing for 28 days at 80 °C in 1M NaOH. It can be seen from the figure that the dolomite crystals inside the rock are relatively complete, and no lamellar brucite and columnar calcite generated by the dedolomization reaction were found, which indicates that the addition of 50% fly ash and 50% metakaolin inhibit the progress of the dedolomization reaction. Figure 10. SEM-EDS image of the product in fracture surface of aggregate S1# in the concrete microbars cured in 1M NaOH solution at 80 °C for 28 days. Figure 10. SEM-EDS image of the product in fracture surface of aggregate S1# in the concrete microbars cured in 1M NaOH solution at 80 • C for 28 days.  Figure 12 shows the expansion cracks of concrete microbars prepared with rock S1# cured in 1M NaOH solution at 80 °C for 56 days. As can be seen from Figure 12a, obvious cracks appeared inside the concrete microbars without fly ash and metakaolin, and these cracks were mainly concentrated in the rock aggregate, from the inside of the rock to the the cement pastes extended, and no ASR gel was found near the aggregate, which indicated that the expansion cracking of the concrete microbars was caused by the ACR. It can be seen from Figure 12b,c that the cracks inside the concrete microbars containing 50% fly ash and 50% metakaolin almost completely disappeared. This indicates that the reduction of expansion cracks of concrete microbars prepared with rock S1# is directly related to the addition of fly ash and metakaolin. Figure 12. The expansion cracks of concrete microbars prepared with rock S1# cured in 1M NaOH Figure 11. SEM image of the product in fracture surface of aggregate S1# in the concrete microbars cured in 1M NaOH solution at 80 • C for 28 days: (a) 50% metakaolin. (b) 50% fly ash. Figure 12 shows the expansion cracks of concrete microbars prepared with rock S1# cured in 1M NaOH solution at 80 • C for 56 days. As can be seen from Figure 12a, obvious cracks appeared inside the concrete microbars without fly ash and metakaolin, and these cracks were mainly concentrated in the rock aggregate, from the inside of the rock to the the cement pastes extended, and no ASR gel was found near the aggregate, which indicated that the expansion cracking of the concrete microbars was caused by the ACR. It can be seen from Figure 12b,c that the cracks inside the concrete microbars containing 50% fly ash and 50% metakaolin almost completely disappeared. This indicates that the reduction of expansion cracks of concrete microbars prepared with rock S1# is directly related to the addition of fly ash and metakaolin.

Effects of Fly Ash and Metakaolin on Expansion Cracks
cracks were mainly concentrated in the rock aggregate, from the inside of the rock to the the cement pastes extended, and no ASR gel was found near the aggregate, which indicated that the expansion cracking of the concrete microbars was caused by the ACR. It can be seen from Figure 12b,c that the cracks inside the concrete microbars containing 50% fly ash and 50% metakaolin almost completely disappeared. This indicates that the reduction of expansion cracks of concrete microbars prepared with rock S1# is directly related to the addition of fly ash and metakaolin. Figure 12. The expansion cracks of concrete microbars prepared with rock S1# cured in 1M NaOH solution at 80 °C for 56 days. (a) control group; (b) 50% fly ash; (c) 50% metakaolin.

Conclusions
The most significant conclusions of this paper are summarized as follows: (1) Four methods were used to determine that the dolomitic limestone S1# is alkali-carbonate reactive aggregate. (2) In the concrete microbars test, metakaolin and fly ash can effectively inhibit the ACR of reactive aggregate, and the more the content of fly ash and metakaolin, the more efficient the effect of inhibition. Less fly ash cannot effectively inhibit the alkali-aggregate reaction of active aggregate. When the content of fly ash and metakaolin is the same, metakaolin is more effective than fly ash.

Conclusions
The most significant conclusions of this paper are summarized as follows: (1) Four methods were used to determine that the dolomitic limestone S1# is alkalicarbonate reactive aggregate. (2) In the concrete microbars test, metakaolin and fly ash can effectively inhibit the ACR of reactive aggregate, and the more the content of fly ash and metakaolin, the more efficient the effect of inhibition. Less fly ash cannot effectively inhibit the alkaliaggregate reaction of active aggregate. When the content of fly ash and metakaolin is the same, metakaolin is more effective than fly ash. (3) The products of ACR calcite and brucite can be clearly observed in the SEM image, but no ASR gel was found, which indicates that the expansion of dolomitic limestone S1# was caused by ACR. (4) Metakaolin and fly ash participate in the secondary hydration of cement, which consumes a large amount of the hydration product Ca(OH) 2 of Portland cement and reduces the alkalinity of the cement hydration product, so the ACR be inhibited. The pore structure of concrete containing fly ash and metakaolin becomes denser, which prevents the diffusion of K + and Na + to the active aggregate.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.