The Risk of Alkali–Carbonate Reaction and the Freeze–Thaw Resistance of Waste Dolomite Slag-Based Concrete

Abstract

The alkali–carbonate reaction (ACR) is a type of alkali–aggregate reaction (AAR) that may lead to serious damage in concrete construction. There is sufficient research on the effect of the ACR on dolomite limestone; however, research on the effect of the ACR on pure dolomite is absent, and there are a large number of dolomite resources that cannot be effectively utilized in civil engineering. This study aims to investigate whether the ACR occurs in pure dolomite spoil and to determine the freeze–thaw resistance of pure waste dolomite slag-based concrete (PWDSC). In this study, X-ray diffraction (XRD) and the lithofacies method (LM) confirmed that the tested samples were pure dolomite. The rock cylinder method (RCM) and rapid preliminary screening testing for carbonate aggregates (AAR-5) were employed to determine the alkali activity of pure dolomite: the RCM indicated a variation of −0.09% in length during the 84-day test period, the AAR-5 exhibited a length expansion rate of 0.03% within 28 days, and the expansion rates were less than 0.1%. These findings suggest that pure waste dolomite slag (PWDS) does not possess alkali activity. The freeze–thaw cycle test showed no significant spalling on the concrete surface, the inside of the cement produced few micro-cracks according to scanning electron microscopy (SEM), and the uniaxial compressive strength (UCS) test showed a decrease of approximately 20% after 200 freeze–thaw cycles. The results verified that ACR does not occur in PWDS and that it can withstand freeze–thaw damage, to a certain extent, when used as concrete coarse aggregate.

1. Introduction

Aggregates are the major constituents of concrete, constituting between 60% and 75% of the concrete volume of Portland cement concrete (PCC), which is the most widely used construction material in the world [1,2,3]. Natural aggregates constitute the majority of concrete volume, and the mining of natural aggregates has significant impacts on the environment [4]. However, a large amount of waste rock slag is produced in some civil engineering projects; this slag can be used as concrete aggregate [5]. Excavated rock slag material is stable and hard enough for some projects, and it is suitable for concrete aggregate, which can reduce resource waste effectively. However, some types of waste rock slag cannot be used as concrete aggregate directly because of AAR, which is known as “concrete cancer” [6,7,8,9]. AAR mainly includes the alkali–carbonate reaction (ACR) and alkali–silica reaction (ASR) [10,11,12]. There is a large amount of research on the ASR [13,14,15,16], while little research exists on the ACR.

As an important component of AAR, ACR may lead to a deterioration of concrete structures, especially dams, bridges, highways, and ports [17]. The first report on ACR was conducted in Canada in 1957 [18], and the concept of ACR was proposed by Gillott in 1963 [19]. The reaction conditions of the ACR can be summarized as a wet environment, alkaline, and alkali-active aggregate, which is consistent with ASR [20,21,22,23]; however, the characteristics are different. ACR presents four characteristics: (1) the damage occurs within a relatively short period and can be significant; (2) concrete expansion due to the ACR primarily manifests in the form of cracking, with some surfaces showing protrusions (in particular, non-prestressed concrete often shows a network or map-like pattern of cracks); (3) under scanning electron microscopy (SEM), ASR gel is not revealed in the reaction product of the ACR; and (4) many mineral additives that are effective in mitigating the ASR do not have the same result for the ACR [24,25]. The ACR can also produce severe deterioration in concrete construction, which is why further research on the ACR is necessary. Although research on the ACR has been conducted for decades, with studies mainly concentrated on dolomitic limestone, there has been little research focus on pure dolomite. Moreover, there are many dolomite resources in southwest China [26]; however, a large proportion of these resources cannot be used efficiently in civil engineering. Therefore, the prospects for the comprehensive development and utilization of dolomite are increasingly broad [27].

There is a lack of consensus regarding the expansion mechanism of the ACR [28]. However, there are two theories: a direct reaction mechanism or an indirect reaction mechanism [29]. The theory of the indirect reaction mechanism holds that dolomitization causes the clay minerals in dolomite to absorb water after exposure, resulting in concrete expansion, while the theory of a direct reaction mechanism holds that dolomitization is the direct cause of expansion [29,30,31]. The dedolomitization reaction is expressed as illustrated in Equation (1) [32]:

$$\begin{array}{ll}\mathrm{CaMg}{\left(\mathrm{C}{\mathrm{O}}_3\right)}_2+2{\mathrm{M}}^{+}+2{\mathrm{O}\mathrm{H}}^{-}\to & {\mathrm{CaCO}}_3+\mathrm{M}\mathrm{g}{\left(\mathrm{O}\mathrm{H}\right)}_2+{\mathrm{C}\mathrm{O}}_3^{2-}+2{\mathrm{M}}^{+}\\ \mathrm{dolomite}& \mathrm{calcite}\hspace{1em} \mathrm{brucite}\end{array}$$

(1)

where M⁺ represents the alkali cations (Li⁺, Na⁺, and K⁺).

In Equation (1), the molar volumes of CaMg(CO₃)₂, Mg(OH)₂, and Ca(CO₃)₂ are 64.3 cm³/mol, 24.6 cm³/mol, and 36.9 cm³/mol, respectively. The total molar volume of the reaction equation is reduced by 4.35%. Based on the aforementioned results, dedolomitization is not the cause of the expansion of the ACR [33]. Research indicates that the reaction conditions of the ACR are strict and limited [34]. Tang [35] suggested that when argillaceous dolomite limestone is present, the ACR can occur with a general clay content of 5% to 20% and approximately equal amounts of dolomite and limestone. Gillott [36] reported that only argillaceous fine-grained dolomite limestone, argillaceous fine-grained limestone dolomite, or argillaceous fine-grained dolomite with specific structural characteristics reacted with alkalis and caused concrete to expand and crack. However, few studies have investigated the ACR in pure dolomite, which is of great importance in determining whether it meets the condition of the concrete alkali–aggregate reaction. Research has concentrated on dolomitic limestone; therefore, it is necessary to study the potential risk of the ACR in pure dolomite to take advantage of pure waste dolomite slag in civil engineering.

On the other hand, in the southwestern region of China, there are numerous cold areas where concrete structures frequently experience freeze–thaw cycles. For this study, pure dolomite from the southwest of China was chosen as the test material. Additionally, concrete is a porous material that allows external moisture to easily penetrate its interior. This continuous exposure to freeze–thaw cycles causes the water within the cement paste matrix to expand, leading to cracks and spalling in the concrete [37].

Based on the content discussed above, this study primarily focused on testing the risk of the ACR occurring in PWDS by using multiple testing methods, such as LM, RCM, and AAR-5, and observing the interior structure by SEM. Further, to utilize dolomite resources effectively, one needs to process the PWDS into concrete coarse aggregate in civil engineering, which incurs human, physical, and financial resources. At the same time, because the dolomite samples used in this research were produced in the southwest of China, where the climate is very cold in winter in some regions, many concrete constructions suffer from freeze–thaw cycles. Hence, it was important to test the concrete freeze–thaw resistance for the pure waste dolomite slag processed to coarse aggregate.

2. Materials and Methods

2.1. Materials

In this work, the main materials included pure waste dolomite slag, cement, sand, and water. The dolomite used in this study was obtained from Kaili, Guizhou province, which is situated in the southwest of China, and the cement, sand, and water were conducted at Chengdu, China. It had a density of 2.75 g/cm³, a moisture content of 0.04%, a water absorption of 0.43%, and a UCS of 111.36 MPa. The raw dolomite samples are shown in Figure 1. The original dolomite samples were off-white, and the surface of the rock samples was visually observed to be intact. To accurately confirm the material composition of the dolomite, X-ray diffraction (XRD, and the devices was produced by Netherlands) was utilized to test the dolomite samples, as shown in Figure 2. To process the dolomite to powder, the diameter should be less than 200 μm.

The XRD results, in Figure 2, show that the dolomite samples consisted of pure dolomite that did not contain impurities, and the total chemical component was CaMg(CO₃)₂.

PO42.5R-grade ordinary Portland cement, produced in Sichuan, China, was selected for this study. The cement possessed a density of 3.14 g/cm³, an initial setting time of 177 min, a final setting time of 229 min, a UCS of 31.8 MPa for 3 days, and a flexural strength of 5.7 MPa for 3 days, which adheres to the Chinese standard GB 175-2007 [38]. In order to determine the effective alkali content, it was necessary to confirm the contents of Na₂O and K₂O. Based on the above, an X-ray fluorescence spectrometer (XRF) was used to test the chemical composition. The results are presented in

Table 1. The chemical components of cement determined by using XRF.

CaO	SiO2	Al2O3	Fe2O3	TiO2	K2O	Na2O	Other
60.78%	20.49%	4.90%	4.07%	1.38%	0.50%	0.46%	7.42%

.

The XRF determination ascertained the content of the effective alkali of the cement, and the calculated result was 0.8%, using Equation (2):

$${\mathrm{Na}}_2{\mathrm{O}}_{\mathrm{e}}=\mathrm{N}{\mathrm{a}}_2\mathrm{O}+0.658 {\mathrm{K}}_2\mathrm{O}$$

(2)

where Na₂Oe is the effective alkali content, %; Na₂O is the sodium oxide content, %; and K₂O is the potassium oxide content, %.

Natural river sand and tap water from Chengdu, China, were used in this study, and the fineness modulus of the sand was 2.43, which was medium-fine sand.

2.2. Test Method

2.2.1. The Lithofacies Method

The lithofacies method (LM) is the preferred approach for assessing the alkali activity of aggregates [39]. This method involves the utilization of polarizing and scanning electron microscopes to identify the rock types, structure, and mineral composition of the aggregate. This makes it possible to ascertain whether or not the aggregate contains alkali-active minerals, to determine the specific category of these minerals, and to calculate their percentage to evaluate the alkali activity. If no alkali-active minerals are present in the aggregate, it can be deemed suitable for use without further testing. However, if the aggregate contains alkali-reactive minerals, an additional test is necessary [40]. The process of LM is shown in Figure 3.

2.2.2. The Rock Cylinder Method

The rock cylinder method (RCM) primarily assesses the presence of alkali–carbonate reactions in aggregates and is not suitable for detecting alkali–silicate reactions [39]. Dolomite samples from Kaili were processed into rock cylinders with a diameter of 9 mm and a length of 35 mm from three different pure dolomite samples in the same region. Then, three groups and 18 samples in total were processed, which ensured the test results were universal.

The dolomite rock cylinders were subsequently immersed in water. The length of the dolomite rock cylinders was measured every 24 h until the difference between the two measurements was less than 0.02%. Once this criterion was met, the length of the rock cylinders was considered as the original length. Subsequently, the dolomite rock cylinders were immersed in a 1 M/L NaOH solution, as shown in Figure 4. To prevent the NaOH solution from being exposed to air and evaporating, thus affecting its concentration, the beaker containing the solution was covered with plastic wrap. Finally, the lengths of the rock cylinders at 7 d, 14 d, 21 d, 28 d, 56 d, and 84 d were measured. If the lengths of the dolomite rock cylinders increased by more than 0.1% within the 84 d timeframe, it was concluded that there was a risk of the ACR occurring. The equation for calculating the expansion rate is shown in Equation (3) [39]:

$${\epsilon }_{\mathrm{r}}=\frac{L_f-L_0}{L_0}\times 100\%$$

(3)

where ε_r is length change at test age, %; L₀ is reference length after immersing in water, mm; and L_f is length at test age, mm.

The RCM effectively determines whether the rock sample shows alkali reactivity. According to the specification, if the length expansion rate of the 84-day specimen was less than 0.1%, it implied the absence of alkali activity in the rock samples; otherwise, the rock samples satisfied the condition for an ACR.

The alkali activity of the aggregate is typically assessed by measuring the length of the sample (the measurement accuracy is 0.001 mm). However, to attain a more comprehensive understanding of the overall changes in the sample, this study attempted to measure the sample’s diameter and calculate the volume change before and after the test for the first time. The calculated volume change was then compared with the molar volume change in the reaction equation of dedolomitization. Due to the potential for errors in measuring the diameter of the two ends of the cylindrical sample during processing, this study measured the diameter at both ends as well as the middle section of the sample. Each measurement was consistently performed using the same measuring tool. To ensure accuracy, the final diameter of the sample was determined by averaging the three measurements.

2.2.3. The AAR-5 Method

Based on the AAR-5 method [41], three groups of samples with nine specimens in total were manufactured using a lab mixer. The samples measured 160 mm in length, with a cross-section measurement of 40 mm × 40 mm, as illustrated in Figure 5. The AAR-5 method employs a single particle size of 5–10 mm aggregate. The effective alkali content of cement was calculated to be 0.8% using Equation (2), and KOH was added to adjust the alkali content of the cement to 1.5%. The composition of the specimens included a cement-to-aggregate ratio of 1:1, with a water-to-cement ratio of 0.30. Once the specimens were formed, they were first cured in a 20 °C ± 2 °C curing room with a relative humidity of 85% for 24 h after demolding. Subsequently, the specimens were precured in a 1 M/L NaOH solution, which was heated to 80 °C for 4 h to ensure that both the internal and external temperatures of the specimens reached 80 °C. The initial length of each specimen was promptly measured and then placed back into the 80 °C NaOH solution with a concentration of 1 M/L.

The average expansion rate of each sample was measured and subsequently utilized as the definitive expansion rate of the mortar bars, and the measurement accuracy of the measuring instrument was 0.001 mm. The expansion rate was regularly measured at 7, 14, 21, and 28 days. Within the 28-day period, when the expansion rate of the specimens was less than 0.1%, it indicated that the aggregate did not possess alkali activity; otherwise, it did. The linear expansion rate was then calculated using Equation (4):

$${\epsilon }_T=\frac{L_T-L_0}{L_0-2\Delta }\times 100\%$$

(4)

where ε_T is the linear expansion ratio of mortar bar, %; L₀ is the measurement of the specimen before subjection to sodium hydroxide solution, mm; L_T is the reading obtained at each period of storage in sodium hydroxide solution, mm; and Δ is the length of test copper nail, mm.

2.2.4. The Method and Mix Design of Freeze–Thaw

This study utilized the slow freezing method to assess if the frost resistance of PWDSC adheres to the Chinese standard GB/T 50082-2009 [42]. The slow freezing method [43] involved placing the specimen in a standard curing environment for 2–3 days, followed by immersion in water at 18–22 °C for 4 days before initiating freezing and thawing. Freezing temperatures were controlled within the range of −20 to −18 °C, while the melting temperatures ranged from 18 to 20 °C. Each cycle included reducing the temperature to freezing, followed by thawing and then returning to freezing temperature. The duration of freezing and thawing in each cycle was no less than 4 h. After completing each freeze–thaw cycle, the specimen was immediately immersed in water at 18–20 °C until the concrete was completely defrosted, and measurements were obtained.

Due to the limited research on the freeze–thaw properties of PWDSC, this study produced six specimens with dimensions of 100 × 100 × 100 mm³ to measure the UCS and resistance to freeze–thaw. Three specimens underwent freeze–thaw cycles, while the remaining specimens were stored in a water tank under a constant air temperature with a hydraulic head of no less than 30 mm as the comparative group [44]. The freeze–thaw method results showed that the frost resistance of PWDSC adheres to Chinese standard GB/T 50082-2009 (similar to ASTM C666/C666M-03) [42,45,46].

For each test, the presented results represent the average of the values obtained from three specimens under identical conditions. The mixture proportioning of PWDSC was based on C30 grade concrete, as shown in

Table 2. Mixture proportioning of PWDSC in kg/m³.

Cement (kg)	Water (kg)	The Pure Dolomite Coarse Aggregate (kg)			Sand (kg)
		16~13.2 (mm)	13.2~9.5 (mm)	9.5~4.75 (mm)	4.75~0.15 (mm)
536.56	241.45	110.00	440.28	550.35	440.28

.

Two hundred freeze–thaw cycles were performed by using a GWGS-40 programmable constant temperature and humidity test chamber.

3. Results and Discussion

This section primarily analyzes the results of the LM, RCM, and AAR-5 test methods used to determine the alkali activity of pure dolomite. At the same time, the freeze–thaw resistance of pure waste dolomite coarse aggregate concrete was also determined.

3.1. The Analysis of Alkali Activity of PWDS

3.1.1. Analysis Using the LM

As shown in Figure 6a, the samples of dolomite were analyzed using the LM, which revealed the presence of residual bacterial and algal structures. The dolomite predominantly consisted of powdery material with a small number of microcrystals. Additionally, the dolomite showed a semi-eukaryotic granular mosaic with irregular bacterial and algal residues. Furthermore, the irregular cavities were filled with microgranular and euhedral granular sparry dolomite. The intergranular pores typically ranged from 0.05 to 0.25 mm, with a maximum size of 1.25 mm, as illustrated in Figure 6b. Intergranular pores and dolomite rings were present, as depicted in Figure 6c. Only one micro-crack (0.01–0.02 mm) was filled with clay, as shown in Figure 6d.

The composition of dolomite was analyzed using the LM, as shown in

Table 3. The composition of dolomite.

Calcite	Dolomite	Clay	Pitch	Gypsum	Pyrite	Quartz	Cryptocrystalline Silicon
0	99%	1%	0	0	0	0	0

.

Based on the LM results, it is evident that the dolomite composition analyzed contained 99% CaMg(CO₃)₂, with 1% clay only present in a single micro-crack. This finding was further supported by the XRD analysis results, which confirmed the dolomite’s purity. At the same time, because of the existence of euhedral granular sparry dolomite, the samples might have alkali activity [47,48]; hence, further investigation was required using RCM and AAR-5 analyses.

3.1.2. Analysis Using the RCM

To better visualize the variation in the dolomite rock cylinders when immersed in NaOH solution, the expansion curve of the test samples is displayed in Figure 7.

In Figure 7, it is evident that the rate of increase in the dolomite rock cylinders was less than 0.1% during the 84-day testing period. The length of all the dolomite rock cylinders ultimately decreased, with an average expansion rate of −0.09%, which indicated that the PWDS did not undergo an alkali–carbonate reaction [49]. However, due to the reduced length of dolomite samples, which satisfied Equation (1) [32], we used SEM to determine the reason for the length reduction in the rock cylinders.

A section of the dolomite sample used in the test was chosen as the subject of investigation and scanned using SEM, which as shown in Figure 8. Due to the cylindrical shape of the sample, both the edge and center positions were selected for study using SEM, as shown in Figure 9.

The result of the SEM is shown in Figure 9, where it is evident that the dolostone edge displayed a more pronounced reaction, as depicted in Figure 9a. Brucite and calcium carbonate formed; however, the reaction did not fully complete. Conversely, there was little observable reaction at the center of the sample, and the dolomite crystal morphology remained unchanged, as depicted in Figure 9b. These results indicate a significantly higher reaction degree at the edge compared with the center.

To quantify the extent of the dedolomitization reaction, we dried the samples before and after the test at 105 °C, measured their volume, and then determined the degree of dedolomitization through the volume change rate. Additionally, by considering the alteration in molar volume in the dedolomitization reaction equation, the final reaction degree can be calculated using Equation (5):

$${\epsilon }_{\mathrm{v}}=\frac{V_f-V_i}{V_i}\times 100\%$$

(5)

where ε_v is the dolomite volume reduction rate, %; V_i is the initial volume of the dolomite rock cylinders, mm³; and V_f is the final volume of the dolomite rock cylinders, mm³.

By analyzing the test data presented in

Table 4. The date of dolomite rock cylinders for RCM.

The First Group	K1-1-1	K1-1-2	K1-1-3	K1-2-1	K1-2-2	K1-2-3	Average Value
The initial diameter (mm)	8.90	8.92	8.92	8.92	8.95	8.93	8.92
The initial length (mm)	35.05	35.11	35.06	35.06	35.09	35.08	35.07
The final diameter (mm)	8.89	8.91	8.91	8.89	8.94	8.93	8.91
The final length (mm)	35.02	35.07	35.04	35.04	35.05	35.04	35.04
The Second Group	K2-1-1	K2-1-2	K2-1-3	K2-2-1	K2-2-2	K2-2-3	Average Value
The initial diameter (mm)	8.93	8.88	8.96	8.96	8.88	8.93	8.93
The initial length (mm)	35.08	35.11	35.06	35.06	35.05	35.08	35.07
The final diameter (mm)	8.91	8.87	8.96	8.96	8.88	8.93	8.92
The final length (mm)	35.04	35.08	35.04	35.03	35.02	35.04	35.04
The Third Group	K3-1-1	K3-1-2	K3-1-3	K3-2-1	K3-2-2	K3-2-3	Average Value
The initial diameter (mm)	8.91	8.97	8.96	8.92	8.89	8.81	8.91
The initial length (mm)	35.07	35.08	35.08	35.04	35.06	35.08	35.07
The final diameter (mm)	8.91	8.96	8.96	8.91	8.89	8.81	8.91
The final length (mm)	35.04	35.05	35.05	35.02	35.03	35.04	35.04

, the average volume reduction rate was determined for each group of samples. The final volume reduction rate of the sample was then obtained by averaging the values from the three groups of data. Additionally, the final volume change rate was computed using Equation (5), resulting in a rate of −0.23%. As previously mentioned, dedolomitization led to a decrease in the molar volume by 4.35%. To calculate the final reaction degree, the volume reduction rate of the sample was divided by 4.35%, and the degree of reaction was divided by 5.3%.

Combining the RCM, SEM, and reaction degree data revealed that, when subjected to the rock cylinder method, the PWDS underwent a dedolomitization reaction with a low degree of reaction, and the dedolomitization concentrated on the edge of the samples. The absence of expansion in the sample indicates the absence of the risk of the alkali–carbonate reaction occurring in pure dolomite. Additionally, these findings indirectly demonstrate that the alkali–carbonate reaction exclusively occurs in dolomitic limestone concrete [50].

3.1.3. The Analysis of AAR-5

• The length variation of the samples

The length variation of the samples, according to AAR-5, is shown in Figure 10.

Successive measurements were obtained after 0, 7, 14, 21, and 28 days using the AAR-5 test method, and the expansion rate of the mortar bars is shown in Figure 10, where it can be observed that the maximum expansion of the mortar bars was 0.04%, the minimum expansion was 0.02%, and the average expansion was 0.03%, which was lower than 0.1%.

2.

The micro-structure of the samples

The result of the AAR-5 method indicated that the alkali–carbonate reaction did not occur in the mortar bars; however, the length of the mortar bars showed a slight expansion. To investigate the cause, this study employed scanning electron microscopy (SEM) to investigate the occurrence of the dedolomitization reaction in the transition zone between the PWDS and mortar using the AAR-5 method to analyze whether dedolomitization would lead to an expansion of the mortar bars. The test area is presented in Figure 11, which is marked by the black line, and the SEM image is presented in Figure 12.

In Figure 12, it is clear that the utilization of the AAR-5 did not result in a significant dedolomitization reaction between the pure dolomite aggregate and mortar. Only a small amount of calcium carbonate and brucite was produced, and the dedolomitization product was primarily concentrated on the surface of the pure dolomite aggregate. The matrix area did not show any calcium carbonate or brucite. Additionally, the transition zone structure within the mortar bar remained intact, with minimal formation of micro-cracks. Only one micro-crack occurred at the junction of the mortar and aggregate, along with a small number of micro-pores. This study suggests that the expansion of the mortar bars is due to the penetration of water from the sodium hydroxide solution into the micro-cracks and micro-pores during the test [51].

Based these findings above, we suggest that PWDS can be used as a concrete aggregate without posing a risk of the alkali–carbonate reaction (ACR). We emphasize that a single test method cannot assess the alkali activity of PWDS effectively; multiple test methods need be used, such as the LM, RCM, and AAR-5 test methods.

3.2. The Freeze–Thaw Resistance of PWDSC

3.2.1. The Surface Characteristics of the PWDSC

Firstly, the surface characteristics of concrete after every 25 freeze–thaw cycles are shown in Figure 13.

In Figure 14, it can be observed that as the number of freeze–thaw cycles increased from 0 to 200, the surface of the dolomite coarse aggregate concrete remained intact, without any noticeable spalling. The initial surface holes of the concrete did not exhibit significant freeze–thaw deterioration.

3.2.2. The Variation in the PWDSC Mass

The mass of concrete was recorded for 0, 25, …, and 200 cycles. To effectively observe the rate of change in mass of the PWDSC, the variation in mass curves is shown in Figure 14.

It can be observed that the mass of the samples increased gradually. However, the increment in the mass was minimal, as depicted in Figure 14. The maximum increase was 0.24%, the minimum increase was 0.22%, and the average increase was 0.23%. The reason for the mass increase in the PWDSC is due to the freeze–thaw cycles causing the originally closed pores to connect through the freeze–thaw attack, which allows water to enter the pores. Subsequently, the water expands upon freezing, further enlarging the previously closed pores and opening more pores.

3.2.3. The UCS of the PWDSC

After 200 freeze–thaw cycles, the UCS test was performed. The UCS of concrete was measured following GB/T 50081–2019 [52]. The applied loading speed was set as 4000 N/S, and the stress–strain curves are shown in Figure 15, and the dotted line represented the samples suffer freeze-thaw destroy, and the solid line represented the samples didn’t suffer freeze-thaw destroy.

$$\Delta {\sigma }_c=\frac{{\sigma }_{c0}-{\sigma }_{cn}}{{\sigma }_{c0}}\times 100$$

(6)

In Equation (6) above, Δσ_c is the UCS loss rate of specimens after N times of freeze–thaw cycles (%); σ_c0 is the UCS value of the control group (MPa); and σ_cn is the UCS value of specimens after N times of freeze–thaw cycles (MPa).

In Figure 15, it can be observed that the UCS of the concrete showed a significant decrease. The average UCS of the non-freeze–thaw concrete was 41.31 MPa, while the average UCS of the freeze–thaw concrete was 33.01 MPa. This represents a decrease of approximately 20% when calculated using Equation (6).

3.2.4. The Characteristics of the Micro-Structure

This finding suggests that the macro-perspective alone cannot directly explain the decline in the strength of the PWDSC. To further investigate, SEM was used to analyze the micro-structure of the PWDSC after 200 freeze–thaw cycles; meanwhile, concrete that did not suffer the destruction of freeze–thaw was used as the control group. The results of SEM are shown in Figure 16.

Figure 16a represents the mortar section, and Figure 16b represents the transition area for the pure dolomite coarse aggregate concrete after freeze–thaw treatment. Figure 16c,d was used as control group. A comparison between Figure 16a,c reveals that the mortar section of the freeze–thaw concrete shows more micro-cracks compared with the control group, regardless of the number or width of the micro-cracks. Similarly, comparing Figure 16b,d shows that there are evident cracks in the transition zone between the concrete aggregate and the mortar in PWDSC that underwent the freeze–thaw cycle. Additionally, the cementation between the coarse aggregate and the mortar was significantly weaker in the freeze–thaw concrete than in the control group. Furthermore, the concrete aggregate that did not undergo the freeze–thaw cycle was more tightly bonded with the mortar, and the pore structure was patently more intact.

In order to further explain the reason for the reduction in the UCS of the PWDSC, it is necessary to generate statistics on the porosity of the concrete. Based on the SEM images, porosity can be quantitatively estimated by gray value-based image analysis [53,54,55,56,57] with Image J. To obtain a black and white image, the voids/pores before and after freeze–thaw for concrete were identified by pixel brightness, as shown in Figure 17. The total porosity of the observed surface of the sample can be assessed by using Equation (7) [58].

$$\phi =\frac{A_{\mathrm{pore}}}{A}\times 100\%$$

(7)

where: φ is the porosity of concrete; A_pore is the measured area of pores; and A is the measured area of the observed surface.

Based on Equation (7) and Figure 17, the porosity can be calculated by using Image J2, and the result is shown in Figure 18.

Based the result shown in Figure 18, the porosity of PWDSC before freeze–thaw cycles can be analyzed is from 9.24% to 11.36% after 200 freeze–thaw cycles, which reflects that the porosity of concrete increases during the freeze–thaw process. This is because the pore water in concrete will undergo the process of crystallization, melting, and recrystallization during the freeze–thaw cycle, and the resulting crystallization pressure will destroy the primary pores of concrete [59,60], leading to an increase in porosity. At the same time, this also explains the reason for the decrease in concrete strength from the side.

To sum up, it can be concluded that the freeze–thaw cycle causes damage to the PWDSC, especially for the micro-structure. However, since the concrete was prepared according to the strength of C30, and because the strength of the concrete after the freeze–thaw treatment was higher than 30 MPa, the pure dolomite effectively resisted the freeze–thaw damage to some extent.

4. Conclusions

This research showed that PWDS does not possess alkali activity and that PWDSC has resistance to freeze–thaw cycles; the detailed conclusions can be summarized as follows:

(1)

The results of the XRD and LM showed that the dolomite slag consisted of 99% dolomite, which suggested that the dolomite samples were pure dolomite and ruled out the risk of an ASR. Meanwhile, due to the existence of euhedral granular sparry dolomite, the samples could have alkali activity; hence, further methods were used to assess the risk of an ACR.

(2)

To confirm the risk of the ACR for PWDS, RCM and AAR-5 were employed. The results of the RCM showed a reduction in the length of the cylinders and did not show expansion after the test subjects were soaked in a 1M/L NaOH solution for 84 days. The length of all the dolomite rock cylinders ultimately decreased by 0.1%, with an average expansion rate of −0.09%. From a microscopic perspective, SEM analysis revealed that the dolomite cylinders carried out a significant dedolomitization reaction. Furthermore, the decrease in the volume change rate of the dolomite cylinders provided additional evidence supporting the reduction in the molar volume associated with the dedolomitization reaction. Additionally, the RCM results corroborated that pure dolomite lacked the necessary conditions for the ACR; the result of the AAR-5 method showed a length expansion rate significantly lower than 0.1%. This observation supported the assertion that PWDS does not possess alkali activity. Additionally, the results provided evidence that the alkali–carbonate reaction was not primarily caused by the reaction between CaMg(CO₃)₂ and alkalis. Simultaneously, a subtle expansion took place in the mortar bars because the saturation effect during the test led to a minor expansion of the mortar bars. SEM revealed the presence of a limited number of micro-cracks and micro-pores within the mortar bars.

(3)

After 200 freeze–thaw cycles, there was no obvious spalling on the surface of the pure dolomite coarse aggregate concrete; however, the UCS decreased by 20%. The porosity of the PWDSC before the freeze–thaw cycles was 9.24%, and it increased to 11.36% after 200 freeze–thaw cycles, which shows that the porosity of the PWDSC increased during the freeze–thaw process, explaining why the UCS declined. Because the concrete strength was in accordance with C30 grade concrete, the strength was still higher than 30 MPa after freeze–thaw treatment, indicating that PWDSC can effectively resist freeze–thaw damage.