Effects of Freeze–Thaw Cycles on Performance and Microstructure of Cold Recycled Mixtures with Asphalt Emulsion

Abstract

Although it is widely recognized that freeze–thaw cycles have a great influence on the properties of asphalt pavement, a quantitative understanding of how freeze–thaw cycles affect cold recycled mixtures with asphalt emulsion (CRME) is so far still lacking. The main objective of the paper was to investigate the performance and microstructure of CRME under freeze–thaw cycles with different water saturation conditions. For this, air voids, high-temperature stability, low-temperature cracking resistance, and moisture susceptibility of CRME were analyzed based on laboratory tests. The micro-morphology and chemical composition of cement asphalt emulsified compound mortar were observed by scanning electron microscopy (SEM). Results showed air voids of CRME increase as freeze–thaw cycles increase; the high-temperature stability, low-temperature cracking resistance, and moisture susceptibility of CRME decrease as freeze–thaw cycles increase; the asphalt strips from the surface of hydration products, and the composite structure mainly consists of hydration products as freeze–thaw cycles increase; the microstructure of CRME is destroyed. The freeze–thaw cycles have a negative effect on the CRME performance and microstructure.

1. Introduction

A large number of asphalt materials are disposed of in China annually, resulting in extra land used for dumping and environmental pollution. Meanwhile, the lack of high-quality aggregates crucially restricts road construction and maintenance, which results in resource and environmental issues [1]. Cold-recycled technology is an efficient method of recycling wasted asphalt materials [2]. Cold-recycled technology has been widely applied in the pavement structure [3,4], which can recycle a massive amount of wasted asphalt materials to decrease environmental pollution [5].

Freeze–thaw cycles lead to a decrease in the durability of asphalt pavement in areas that experience seasonal freezing. The destructive effects of freeze–thaw cycles are caused by water expansion in the asphalt mixture when the temperature is lower than 0 °C. The movement and phase transition of water changes the internal characteristics of the asphalt mixture [6,7,8]. CRME features plenty of air voids, meaning that water could enter the CRME interior easily; therefore, the freeze–thaw damage to CRME is worse than a hot mix asphalt.

A significant amount of research has been conducted to study the performance damage of asphalt mixtures under freeze–thaw cycles. Feng D. et al. assessed the tensile strengths of asphalt mixtures under salt and freeze–thaw cycle conditions. Results showed that salt had a remarkable influence on tensile strength with more than 3% salt; freeze–thaw cycles showed a larger effect when there was less than 3% salt [9]. Lachance-Tremblay et al. assessed the moisture susceptibility and degradation of a glass asphalt mixture through complex modulus under freeze–thaw cycles. Results showed that the mixtures were destroyed under freeze–thaw cycles and destruction was more severe than the normal mixture experienced in 10 freeze–thaw cycles [10]. You L. et al. evaluated the damage degree of water-foamed asphalt mixture with multiple freeze–thaw cycles. The results indicated that freeze–thaw cycles accelerated the internal damage of the asphalt mixture and resulted in an increase in air voids and a decrease in adhesive strength [11]. Fan Z. et al. investigated the influence of freeze–thaw cycles on the fatigue performance of asphalt mixture. The results showed that fatigue life decreased as saturation and freeze–thaw cycles increased. The fatigue life decreased significantly at high stress levels [12]. Duojie C. et al. illustrated the mixtures’ flexural tensile properties via a bending test subjected to freeze–thaw conditions. Results showed that the flexural tensile strength and strain decreased with the increase in freeze–thaw cycles. The deterioration of flexural tensile properties decreased severely at the beginning of the freeze–thaw cycles but became smooth after the freeze–thaw cycles [13]. Huang T. et al. investigated the effects of freeze–thaw cycles on AC-13 and SMA-13 asphalt mixtures. It was found that the multi-axial strength obviously decreased when subjected to 20 freeze–thaw cycles. The SMA-13 asphalt mixture demonstrated good freeze–thaw resistance [14]. Fu L et al. studied the influence of freeze–thaw cycles and aging on the damage properties of asphalt mixture subjected to splitting loads. The findings demonstrate that the freeze–thaw cycles and aging effects affected the damage properties of asphalt mixtures, resulting in early damage, exacerbating the formation of micro-cracks in the early stage, accelerating the expansion of macro-cracks, and advancing the shedding between the asphalt and aggregates [15]. Dan H.C. et al. investigated the influence of freeze–thaw cycles on the mechanical properties, thermal conductivity, and water permeability of asphalt mixtures in the cold region. The findings demonstrated that the mechanical properties and thermal conductivity of asphalt mixtures decayed with the increase in freeze–thaw cycles, although the water permeability of asphalt mixtures was improved [16]. Chai C. et al. focused on the resistance of porous asphalt mixture in various air voids to freeze–thaw cycles in order to study the suitable gradation for seasonal freezing regions. The results showed that, with the increase in air voids, the indirect tensile strength and compression strength decreased, and the particle loss resistance and high-temperature stability worsened [17]. El-Hakim et al. found the dynamic modulus results did not show a statistically significant difference between the average |E*| of the SP 25 and SP 25 RBM after construction; however, the benefits of additional binder content were apparent after simulating freeze–thaw cycles over 1 year in service [18]. Ud Din et al. found freezing and thawing could severely affect the compressive strength, air voids, fatigue cracking, and the rutting of the asphalt pavement. Fatigue and rutting were distresses associated with the flexible pavement that is more sensitive to climatic conditions [19]. Guo Q. et al. investigated the interface properties under the impacts of snowmelt salt and freeze-thaw cycles. The results indicated that the pull-off tensile strength loss after salt solution soaking and freeze-thaw cycles were all higher than the pure water soaking one. Adhesive failure was prone to occur at low temperature under the conditions of low salt concentration and freeze-thaw cycles. The erosion effect of a high concentration solution would induce the adhesive failure at a medium temperature [20].

The performance of the asphalt mixture is related to the internal void characteristics. Xu H. et al. obtained the internal void information of different kinds of asphalt mixtures under freeze–thaw cycles by X-CT. The results indicated there were three internal void evolution behaviors: (1) enlargement of original voids; (2) connection of independent voids; (3) appearance of new voids [6], demonstrating the non-negligible effect of pore structure on dynamic flow evolution under freeze–thaw cycles [21]. In addition, Xu H. et al. explained the internal structural properties of asphalt mixture through the information entropy theory under freeze–thaw cycles [22]. Gong Y. et al. studied the air voids and morphological characteristics of nano-TiO₂/CaCO₃ and basalt-fiber-modified asphalt mixtures subjected to freeze–thaw cycles. It was found that the macro voids began to connect gradually under 15 freeze–thaw cycles. The shape and number of air voids were altered [23]. Gao X. et al. characterized the three-dimensional internal structure evolution of asphalt mixtures subjected to freeze–thaw cycles. The results showed asphalt mixture void ratio increased with the increase in freeze–thaw cycles; the larger the initial air voids, the more severe the increase in the samples [24].

Though there are numerous studies on hot-mixed asphalt mixtures available in the literature, very few have studied the influence of freeze–thaw cycles on CRME, even though it is of great importance to CRME application in seasonally frozen areas. Previous studies of the CRME mainly focused on mix design, performance evaluation, early-age strength, additives, and so on. Xu O. et al. assessed the characteristics of the cement emulsified asphalt mixtures from the effects of aggregate gradations and binder contents. The results showed that high-temperature stability could be promoted through the suitable aggregate fractal feature and cement content. Aggregate gradation and binder content had greatly impacted the moisture susceptibility and indirect tensile strength. The Cantabro loss index was mainly affected by aggregate gradation [25]. Kim Y. et al. adopted raveling tests, dynamic modulus, flow time, and flow number to evaluate the short-term and long-term performance of CIR-emulsion mixtures [26]. Gao L. et al. simulated five levels of mixed-mode cracking by a newly developed Arcan configuration [27]. Jiang J. et al. investigated the high-temperature performance of CRME based on both binder and mixture laboratory tests. Experiment results showed that the high-temperature performance of CRME could be improved significantly by adding 6% styrene-ethylene-butylene-styrene, 8% chloroprene rubber latex, and 6% styrene-butadiene rubber latex in the asphalt emulsion [28]. Zhang J. et al. investigated the effects of 10 factors on the performance of CRME subjected to the low-temperature anti-cracking based on the semicircular bending test with fracture energy as the evaluation index. It was found that the biggest factor was the recycling agent, followed by cement, fiber, and compaction work [29]. Yan J. et al. found that cement played a positive role in early-age strength and long-term performance of cold-recycled mixes [30]. Chen T. et al. suggested that cement and asphalt emulsion should be fully mixed before the contact between aggregate and mortar [31]. Yang Y. et al. evaluated the impact of cement on the performance of CRME with various cement contents and the complex interaction between asphalt emulsion and cement. It was recommended that cement content ranges between 1% and 2% [32]. Dong S. et al. evaluated the effects of different modification approaches on the mechanical properties and durability of CRME. Results showed that the comprehensive performance of the modified CRME could be improved by the adjunction of styrene–butadiene rubber latex, Buton rock asphalt, and rejuvenation agent. Among these additives, the rejuvenation agent had the most significant effect, followed by Buton rock asphalt [33]. Du S. applied polyester fiber, polypropylene fiber, polyacrylonitrile fiber, lignin fiber, and basalt fiber to the emulsion recycled mixture. The results indicated that polyester fiber should be recommended as the most effective fiber to improve the emulsion recycled mixture [34]. Yang Y. et al. investigated the meso-structural characteristics of CRME. The results showed the evolution of CRME could be divided into three stages: the stage of initial compaction, relative compaction, and destruction [35]. So far, a quantitative understanding of how freeze–thaw cycles affect CRME is still lacking.

The main objective of the paper was to investigate the performance and microstructure of CRME subjected to freeze–thaw cycles with different water saturation conditions. Due to the large void ratio (8–13%) of CRME, water could easily enter the internal mixture and reduce the cohesion between asphalt and aggregate. The resistance performance of the erosion from high-pressure water repeatedly will decrease. The high-temperature deformation of CRME will reduce the smoothness of pavement. The low-temperature cracking will cause significant pavement distress. Air voids, road performance subject to temperature, and CRME water content should be investigated. SEM was applied to observe the microstructure of fracture surfaces and measure the chemical composition of mortar on CRME. The damage mechanism of CRME was analyzed under freeze–thaw cycles. This study will be significant to knowing the effects of freeze–thaw cycles on, and promoting the development of, CRME in the seasonal frozen area.

2. Materials and Methods

2.1. Raw Materials

The cationic slow-cracking asphalt emulsion was adopted in the study, whose characteristics are shown in

Table 1. Characteristics of asphalt emulsion.

Characteristic	Requirements	Results
Remained content on 1.18 mm/wt.%	≯0.1	0.021
Solid content/wt.%	>60	63.6
Penetration (25 °C, 100 g)/0.1 mm	50–130	69.5
Softening point/°C	-	45.6
Ductility (15 °C)/cm	≮40	76.5
Solubility in trichloroethylene/wt.%	≮97.5	99.1
Storage stability at 1 d/wt.%	≯1	0.4
Storage stability at 5 d/wt.%	≯5	2.6

, according to JTG T5521-2019 [36]. RAP was obtained from Shenying highway in Liaoning, China. In order to avoid the influence of RAP gradation, the milling RAP in the pavement was divided into 0–0.075 mm, 0.075–0.15 mm, 0.15–0.3 mm, 0.3–0.6 mm, 0.6–1.18 mm, 1.18–2.36 mm, 2.36–4.75 mm, 4.75–9.5 mm, 9.5–13.2 mm, 13.2–16 mm, and 16–19. Meanwhile, the unreal coarse aggregate, which was made up of fine aggregate bonding, was eliminated during the screening process. The sand content of RAP was 65, which was more than the requirement of 50. The new aggregate was joined to ensure the performance of CRME. The requirements of the new aggregate are shown in

Table 2. Characteristics of coarse aggregate.

Characteristic	Requirements	Results
Crushed value/wt.%	≯26	11.7
Los Angeles abrasion/wt.%	≯28	8.8
Flat and elongated particles/wt.%	≯15	5.1
Water absorption/wt.%	≯2.0	0.45
Particle size <0.075 mm/wt.%	≯1	0.3

and

Table 3. Characteristics of fine aggregate.

Characteristic	Requirements	Results
Sand content/wt.	≮60	79.3
Angularity/s	≮30	44.3
Methylene blue value/(g/kg)	≯25	1.7

, according to JTG F40-2004 [37]. The new aggregate was divided into a single size to avoid the influence of graded variation. The biggest size of RAP was less than the nominal maximum aggregate size of 19.0 mm according to JTG T5521-2019 in China [36]; therefore, a new aggregate of 19–26.5 mm was added. In addition, 32.5# of ordinary Portland cement was added to improve the early-age strength and performance of CRME [32]. Drinking water free of impurities was applied as a pre-mix water to improve the coating of the RAP and virgin aggregates, lubricate the mix during compaction, and accelerate the cement hydration reaction.

2.2. Mixture Design and Sample Preparation

2.2.1. Mixture Design

The nominal maximum aggregate size of CRME was 19 mm according to JTG T5521-2019 [36]. The CRME consisted of RAP (68.95 wt.%), new aggregate (29.55 wt.%), and ordinary Portland cement (1.5 wt.%). The design gradation of RAP and new aggregate is shown in

Table 4. Design gradations of RAP and new aggregate.

Size/mm	Passing Rate %
	RAP	New Aggregate
26.5	100	100
19	100	93.9
16	88.8
13.2	78.7
9.5	63.5
4.75	38.1
2.36	22.8
1.18	13.7
0.6	7.6
0.3	4.1
0.15	2.0
0.075	1.0

. The content of cement was determined based on previous research [28]. The graduation curve of CRME is illustrated in Figure 1. The design process of CRME was based on JTG T5521-2019 [36]. Firstly, the optimum water content of CRME was determined according to JTG E40-2007 [38]. Secondly, the optimal asphalt emulsion content of CRME was defined through air voids, split strength, and dry–wet splitting strength ratio under the optimum water content according to JTG T5521-2019 [36]. Finally, the optimal content of asphalt emulsion and water of CRME were 3.5 wt.% and 3.0 wt.%, respectively.

2.2.2. Sample Preparation

All test samples were prepared using the Superpave gyratory compactor with an angle of gyration of 1.25°, vertical pressure stress of 0.6 MPa, and gyration speed of 30 rpm. All compacted samples were 100 ± 0.5 mm in diameter and 63.5 ± 1.3 mm in height. The CRME samples were compacted at an optimal dosage of water and asphalt emulsion. At first, the CRME was placed in the mold and compacted until the height of the sample was 63.5 mm. Secondly, the whole compacted samples without mold were placed in a draft oven at 60 °C for at least 48 h. Finally, the cured samples were placed at room temperature for at least 12 h.

2.3. Experimental Methods

2.3.1. Freeze–Thaw Cycle Tests

Freeze–thaw cycle tests with various water saturation conditions were adopted after curing. Different water saturation conditions were: dry condition, half water saturation condition, and complete water saturation condition.

(1) The 0% water saturation condition was a dry condition. The surfaces of all the samples were coated with plastic wrap after curing.

(2) The 100% water saturation condition was a complete water saturation condition. Firstly, all the samples were conditioned in the water by vacuum saturation using a residual pressure of 98.3–98.7 kPa for 15 min. Then, all samples were conditioned in the water for at least 2 h at normal pressure to constant weight. Finally, the surface of the samples was dried with a soft cloth and then coated with plastic wrap.

(3) The 50% water saturation condition was the half water saturation condition. All samples with the complete water saturation condition were weighed to calculate the added weight. After that, the whole samples were put into a draft oven to reduce half of the added weight. Finally, their surfaces were coated with plastic wrap and then left at room temperature for at least 12 h.

All whole samples were frozen in the freeze–thaw cycles test box after preparation. The samples were frozen at −20 °C for 6 h and then thawed at 20 °C for 6 h [13,39,40]. This process constitutes a complete freeze–thaw cycle. The specimens were subjected to 0, 5, 10, 15, and 20 cycles, respectively. After the freeze–thaw cycles, the whole samples were dried to measure the properties of CRME. Figure 2 illustrates the test samples before testing.

2.3.2. Air Voids Test

Air void calculations needed the relative bulk density and theoretical maximum relative density of CRME. The relative bulk density of CRME was obtained by the density of compacted asphalt mixture test (dry surface method) in accordance with JTG E20-2011. The relative bulk density was calculated by Equation (1).

Υ_f = m_a/(m_f − m_w)

(1)

where Υ_f, m_a, m_f, and m_w represent the relative bulk density (dimensionless), the air weight of the dried sample (g), the surface dry weight of the sample (g), and the weight of the sample in water (g), respectively.

The theoretical maximum relative density of CRME was obtained by the theoretical maximum relative density test (vacuum method) in accordance with JTG E20-2011. The theoretical maximum relative density of CRME was calculated by Equation (2).

Υ_t = m_b/[m_b − (m₂ − m₁)]

(2)

where Υ_t, m_b, m₁, and m₂ represent the theoretical maximum relative density (dimensionless), the air weight of dried CRME (g), the weight of the negative pressure-vessel in 25 °C water (g), and the weight of the negative pressure-vessel and CRME in 25 °C water (g), respectively. Air voids were calculated by Equation (3).

VV = (1 − Υ_f/Υ_t) × 100

(3)

where VV represents air voids (%), respectively.

2.3.3. High-Temperature Stability Test

The high-temperature stability of CRME was evaluated based on the uniaxial penetration test. The size of the head was 50 mm × 50 mm × 10 mm. The size of the bottom was φ28.5 mm × 50 mm. The uniaxial penetration test was performed at 60 ± 1 °C with a loading speed of 1 mm/min based on JTG D50-2017 [41]. A universal electromechanical tester (70-S18B2) (Controls S.R.L, Milano, Italy) was adopted to perform the uniaxial penetration test to measure the maximum load and penetration depth. The maximum load was regarded as the failure strength of the CRME sample. The shear strength was calculated by Equation (4).

R_S = f P/A

(4)

where R_S, f, P, and A represent the shear strength (MPa), the sample dimension correction coefficient f = 0.34, the maximum load (n), and the cross-sectional area (mm²), respectively.

2.3.4. Low-Temperature Cracking Resistance Test

The low-temperature cracking resistance of CRME was evaluated based on the indirect tensile strength (IDT) test, which was conducted at −10 ± 0.5 °C with a loading speed of 1 mm/min according to JTG E20-2011 [42]. This loading generates a relatively uniform tensile stress perpendicular to the direction of the applied load and along the vertical diametrical plane, which causes the specimen to fail by splitting along the central part of the vertical diameter. A universal electromechanical tester (70-S18B2) (Controls S.R.L, Milano, Italy) was applied to perform the IDT test to the ultimate load. The IDT was calculated by Equation (5).

R_T = 0.006287F/h

(5)

where R_T, F and h represent the IDT (MPa), the ultimate load (n), and the height of the CRME sample (mm), respectively.

2.3.5. Moisture Susceptibility Test

The moisture damage of CRME was evaluated based on the freeze–thaw spit test, which was subjected to 25 ± 0.5 °C with a loading speed of 50 mm/min according to JTG E20-2011 [42]. A universal electromechanical tester (70-S18B2) (Controls S.R.L, Milano, Italy) was applied to perform the IDT test to the ultimate load. The IDT was calculated by Equation (5) after different freeze–thaw cycles in different water saturation conditions. The percentage of the IDT ratio, TSR (%), was obtained by Equation (6).

TSR = R_Ti/R_T0 × 100

(6)

where R_T0 represents the IDT of 0 freeze–thaw cycles in different water saturation conditions (MPa), and R_Ti represents the IDT of different freeze–thaw cycles in different water saturation conditions (MPa), respectively.

2.3.6. SEM Test and Energy Spectrum Analysis

The sample was broken by the press machine at a constant rate after different freeze–thaw cycles with different water saturation conditions. About 10–15 mortar slices from the same sample were put into clean containers for observation. The SEM (Hitachi Limited, Tokyo, Japan) and energy-dispersive system by PHILIPS-FEI Quanta 200 were adopted to observe the phase composition and chemical composition of the CRME fracture surface. The energy-dispersive system of SEM PHILIPS-FEI Quanta 200 was applied to analyze the chemical elements greater than Boron (B).

3. Results and Discussion

3.1. Air Voids

As schematically illustrated in Figure 3, the air voids of CRME increased slowly as freeze–thaw cycles increased in dry conditions. The structure of CRME was damaged slightly from the temperature stress in dry conditions. A few disconnected aggregates fell off during temperature cycles. The air voids of CRME only increased up to 3.0% in dry conditions after 20 freeze–thaw cycles; however, the air voids of CRME increased severely as freeze–thaw cycles increased in the non-dry conditions. The structure of CRME was seriously damaged due to the action of temperature stress with water erosion. The internal closed voids were connected as freeze–thaw cycles increased in the non-dry conditions. A large number of new air voids were produced with fine aggregates falling. After 20 freeze–thaw cycles, the air voids increased up to 14.2% and 19.5% in the half water saturation condition and complete water saturation condition, respectively. The action of temperature stress with water erosion was stronger in the complete water saturation condition than in others. The air voids of CRME increased most as freeze–thaw cycles increased in the complete water saturation condition. Due to the large void ratio and structural characteristics of CRME, the increase in air voids was faster than the hot-mix asphalt mixture under freeze–thaw cycles in the complete water saturation condition [11,19,23].

3.2. High-Temperature Stability

As schematically illustrated in Figure 4, the shear strength of CRME decreased as freeze–thaw cycles increased in different water saturation conditions. The shear strength decreased significantly when the water saturation ratio increased. After 20 freeze–thaw cycles, the shear strength decreased up to 20.6%, 31.8%, and 49.1% in dry condition, half water saturation condition, and complete water saturation condition, respectively. The internal damage of CRME occurred due to temperature stress in the dry condition. The damage of CRME accumulated gradually as freeze–thaw cycles increased in the dry condition. The shear strength of CRME decreased in the dry condition. The serious internal damage of CRME occurred due to the temperature stress with water erosion in the half water saturation condition and complete water saturation condition. The cohesion of the mixture decreased. The shear strength of CRME decreased significantly in the half water saturation condition and complete water saturation condition. Compared to the hot-mix asphalt mixture, the decrease in high temperatures was significant under freeze–thaw cycles, which was the result of the rapid increase in air voids and the loss of shear strength [17,19].

As schematically illustrated in Figure 5, the penetration depth of CRME increased as freeze–thaw cycles increased in different water saturation conditions. The penetration depth increased significantly with the water saturation ratio increasing. After 20 freeze–thaw cycles, the penetration depth increased up to 24.5%, 53.5%, and 80.4% in dry condition, half water saturation condition, and complete water saturation condition. Due to the decrease in shear strength, the high-temperature deformation resistance of CRME decreased in the dry condition. The penetration depth of CRME increased gradually as freeze–thaw cycles increased in the dry condition. Due to the significant decrease in shear strength in the half water saturation condition and complete water saturation condition, the high-temperature deformation resistance of CRME decreased significantly. In addition, the air voids significantly increased in the half water saturation condition and complete water saturation condition, which meant that the CRME would be easily compacted under loading. The penetration depth of CRME obviously increased due to the decrease in shear strength and the increase in air voids in the half water saturation condition and the complete water saturation condition.

3.3. Low-Temperature Cracking Resistance

As schematically illustrated in Figure 6, the indirect tensile strength of CRME decreased as freeze–thaw cycles increased in different water saturation conditions. The indirect tensile strength decreased significantly with the water saturation ratio increasing. After 20 freeze–thaw cycles, the indirect tensile strength decreased up to 22.5%, 44.7%, and 56.1% in dry condition, half water saturation condition, and complete water saturation condition, respectively. The internal damage of CRME occurred due to temperature stress in the dry condition. The damage of CRME accumulated gradually as freeze–thaw cycles increased in the dry condition. The indirect tensile strength of CRME decreased in the dry condition. The asphalt may be stripped from the surface of hydration products due to the action temperature stress with water erosion. CRME gradually transformed into the hard-brittle material. In addition, CRME could have a more brittle behavior under low-temperature conditions. The indirect tensile strength of CRME decreased significantly in the half water saturation condition and the complete water saturation condition. The binder of CRME was mainly made up of hydration products as freeze–thaw cycles increased, which was different from the hot-mix asphalt mixture. The CRME was easily damaged in low-temperature conditions. The low-temperature cracking resistance was worse than the hot-mix asphalt mixture [13].

3.4. Moisture Susceptibility Test

As schematically illustrated in Figure 7, the TSR of CRME decreased as freeze–thaw cycles increased in different water saturation conditions. The TSR decreased significantly when the water saturation ratio increased. After 20 freeze–thaw cycles, the TSR decreased by up to 21.5%, 34.9%, and 51.5% in dry condition, half water saturation condition, and complete water saturation condition. Due to the temperature cycle, the fine aggregates fell off and the cohesion of the mixture decreased; therefore, the TSR decreased gradually as freeze–thaw cycles increased in the dry condition; however, the cohesion of the mixture decreased seriously under the action of the coupling of temperature cycle and water. On account of this, the air voids increased significantly as freeze–thaw cycle increased. The TSR of CRME decreased obviously in half water saturation condition and complete water saturation condition. Due to the large void ratio and the rapid increase in air voids, water could easily enter the internal mixture and reduce the cohesion between asphalt and aggregate, which caused the significant decline of moisture susceptibility in comparison to hot-mix asphalt mixture under freeze–thaw cycles in the complete water saturation condition [10,20,21].

3.5. Microstructure Analysis

The microstructure of mortar with different freeze–thaw cycles in the complete water saturation condition is shown in Figure 8. As shown in Figure 8a, hydration products were found on the surface of the mortar. At the same time, hydration products and asphalt were a semi-wrapped condition. The connection between asphalt membrane and hydration products was smooth and continuous, which illustrated that they have good adhesion conditions; however, Figure 8b shows that the mortar was mainly composed of hydration products after 10 freeze–thaw cycles. The asphalt stripped from the surface of hydration products caused the formation of new air voids, coalescing of two separated air voids, and the expansion of existing individual voids. Due to the increase in air voids, the deformation of CRME easily occurred under repeated loading at the high-temperature condition; thus, the high-temperature stability became weak. The cohesion among asphalt and aggregate was poor, so the moisture susceptibility of CRME decreased. Meanwhile, the hydration products became needle-like. CRME could become brittle at low temperatures. As shown in Figure 8c,d, with an increase in freeze–thaw cycles, the asphalt was significantly stripped from the surface of hydration products, and the hydration products became fewer, shorter, and finer. It indicated that the freeze–thaw cycles had a negative effect on the CRME microstructure.

3.6. Energy Spectrum Analysis

The major chemical elements of asphalt emulsion are C, H, S, O, and N. The major chemical elements of the cement are Ca, Si, Al, Fe, and O. The energy spectrum test of the CRME mortar was performed to investigate the change of the crystalline substance of CRME. As schematically illustrated in Figure 9 and Figure 10, the element weight percentage of Si and Ca increased to 25.85% and 11.67%, respectively. The weight percentage of C decreased to 4.18% in the mortar subjected to 20 freeze–thaw cycles in the complete water saturation condition. The results illustrated that the asphalt fell off from the surface of the hydration products and that the mortar mainly included hydration products. It indicated that the composite mortar was damaged by the freeze–thaw cycles.

4. Conclusions

The performance and microstructure of CRME were investigated under freeze–thaw cycles with different water saturation conditions in this paper. Based on the results and discussions, the following conclusions were drawn:

• The air voids of CRME increase as freeze–thaw cycles increase; however, the high-temperature stability, low-temperature cracking resistance, and moisture susceptibility of CRME decrease as freeze–thaw cycles increase. With the increase in saturation rate, the performance changes become more obvious under freeze–thaw cycles.
• The fracture surface of CRME is investigated using the SEM test. The asphalt strips from the surface of hydration products and the hydration products gradually become little, short, and fine as freeze–thaw cycles increase, which illustrates that the freeze–thaw cycles have destroyed the CRME microstructure.
• The chemical composition of CRME is investigated using energy spectrum analysis. The element contents of Si and Ca increased, and the element content of C decreased, which verifies the falling off of asphalt, which also illustrates that the mortar gradually develops into the cement materials.
• According to the analysis of the fracture surface and chemical composition, the asphalt is stripped from the surface of hydration products, which causes an increase in air voids (including the formation of new air voids, coalescing of two separated air voids, and expansion of existing individual voids), and a decrease in cohesion; therefore, the high-temperature stability and moisture susceptibility of CRME decrease. The composite structure mainly consists of hydration products as freeze–thaw cycles increase, resulting in a decrease in the low-temperature cracking resistance of CRME; therefore, we conclude that the freeze–thaw cycles have a negative effect on the CRME performance and microstructure.