Durability and damage evolution of cement-fly ash stabilized aeolian sand gravel under high-temperature curing and freeze–thaw cycles

Abstract

The scarcity of road construction materials in arid desert areas necessitates the utilization of local resources to ensure sustainable infrastructure development. This study investigates the durability and damage evolution of a semi-rigid base material composed of Cement-Fly Ash Stabilized Gravel with 100% aeolian sand replacement for fine aggregates (CFSAG). The mechanical performance was evaluated under simulated desert environments, including high-temperature curing (30 °C, 40 °C, and 50 °C) and freeze–thaw cycles in both water and 2% Na₂SO₄ solution. Additionally, Digital Image Correlation (DIC) technology was employed to characterize the microscopic strain fields and crack propagation patterns. The results indicate that elevated curing temperatures significantly enhance compressive strength, with 40 °C identified as the optimal condition for strength formation due to accelerated hydration. Conversely, resistance to freeze–thaw cycles decreased with higher aeolian sand content, and sulfate erosion accelerated surface spalling, causing the specimen mass to exhibit an "increase-then-decrease" trend. A strength evolution model was established based on freeze–thaw frequency and mix proportions, achieving a fitting accuracy of over 98%. Furthermore, DIC revealed that the strain and displacement variations during crack propagation in specimens exhibited a three-stage evolutionary pattern, with strain localization bands becoming more pronounced as compaction decreases. The study concludes that CFSAG is a feasible, durable, and eco-friendly solution for semi-rigid pavement bases in desert regions.

Introduction

Arid desert regions are characterized by complex natural environments and a scarcity of conventional road construction materials, which poses significant challenges to infrastructure development. However, the discovery of abundant mineral and natural resources in these areas has necessitated the expansion of transportation networks. To address the shortage of construction materials and reduce project costs, the utilization of local aeolian sand resources to replace conventional fine aggregates in the semi-rigid base of asphalt pavements has emerged as a critical research direction^1,2,3,4. Adopting this "locally sourced materials" strategy not only mitigates the durability issues associated with the inferior quality of available materials but also offers substantial economic and environmental benefits by shortening transport distances and conserving resources. Consequently, the feasibility of engineering applications for aeolian sand has garnered extensive attention domestically and internationally^5,6,7,8.

Existing literature has predominantly focused on the application of aeolian sand as fine aggregate in building construction and road engineering. Luo et al.⁹ and Li et al.¹⁰ compiled engineering applications and laboratory tests, confirming that incorporating aeolian sand into cement mortar or concrete is feasible and can improve working properties. Further studies by Amel.¹¹, Hai.¹² , and Al-Harthy et al.¹³ demonstrated that the mechanical properties of aeolian sand concrete are comparable to those of ordinary concrete, even with replacement ratios ranging from 10 to 100%. In the context of road performance, Netterberg et al.¹⁴ observed the stability of aeolian sand subgrades in South Africa, while Amhadi et al.¹⁵ proposed blending manufactured sand with aeolian sand to meet pavement base strength requirements. Additionally, Zhang et al.¹⁶ successfully utilized a mixture of cement, fly ash, gravel, and aeolian sand to reduce construction costs, and Zhai et al.¹⁷ optimized pavement structures using solid waste fly ash to enhance density and durability. Nguyen T A et al.^{18,19,20,21,22} investigated the application of fly ash in building fire-resistant materials through a combined macro- and micro-scale experimental approach. Their research demonstrated that the incorporation of fly ash significantly enhances the material’s mechanical properties and durability, confirming its status as a high-performance solid waste recycling resource. Al-Aghbari et al.²³ further evaluated cement-improved aeolian sand, concluding that such mixtures possess adequate compressive strength and durability for base courses.

Despite these advancements, pavement structures in arid deserts are subjected to extreme environmental conditions, including high temperatures in summer and freeze–thaw cycles coupled with salt erosion in winter, which severely impact their service life. Research by Ruan et al.²⁴ and Yan et al.²⁵ indicated that high-temperature curing conditions, such as 70 °C, can deteriorate stress–strain characteristics and reduce the compressive strength of cement-stabilized aeolian sand. Conversely, Mohammad et al.²⁶ explored the thermal performance of manufactured sand concrete. Regarding freeze–thaw durability, Wu et al.²⁷ noted a lack of comprehensive studies on aeolian sand mixtures under the coupling effects of salt intrusion and freeze–thaw cycles. While Xue et al.²⁸, Zou et al.²⁹, and Dong et al.³⁰ have investigated pore structure evolution and microscopic damage mechanisms in aeolian sand concrete using nuclear magnetic resonance and scanning electron microscopy, research specifically targeting the durability of semi-rigid bases under these harsh desert conditions remains insufficient.

To accurately characterize the damage and deformation behavior of these mixtures, macroscopic mechanical testing alone is inadequate. The Digital Image Correlation (DIC) technique has proven effective in monitoring strain field evolution and crack propagation. Fan et al.³¹, Li et al.³², and Yang et al.³³ successfully utilized DIC to analyze failure characteristics in ultra-high performance concrete, gangue concrete, and expansive soils, respectively. Similarly, Li et al.³⁴ and Chen et al.³⁵ applied DIC to investigate crack evolution regulations and damage variables in concrete. Further applications by Yang et al.³⁶ and Lei et al.³⁷ explored the damage processes in fiber-reinforced concretes, while Wang et al.³⁸ verified the applicability of DIC for lifecycle monitoring of reinforced concrete under corrosive environments. However, the application of DIC to visualize the deformation process and crack extension patterns of cement-fly ash stabilized aeolian sand gravel (CFSAG) bases remains limited.

Current research has revealed several gaps, primarily focusing on three aspects. Understanding remains limited regarding the mechanical properties of aeolian sand mixtures under the double coupling effect of gravel and fly ash. Theoretical studies on the durability of semi-rigid bases under salt-free freeze–thaw cycles are insufficient. There is a lack of multi-scale deformation analysis based on DIC. Therefore, this study systematically established a theoretical framework for a composite system utilising 100% aeolian sand as fine aggregate, specifically the ‘aeolian sand-fly ash’ composite system. By simulating the high-temperature construction environment of desert regions during summer for high-temperature maintenance, the evolution patterns of its mechanical properties were systematically evaluated. Based on different freeze–thaw erosion coupling environments, the durability mechanism of this system under severe conditions was elucidated, thereby filling the theoretical gap in the mechanical properties and durability of CFSAG under extreme environments. In methodology, the innovative introduction of DIC technology enabled full-field, dynamic, and quantitative observation of crack initiation and propagation on the specimen surface during damage progression. This provided a microscale clarification of the damage evolution mechanism. The research results provide crucial theoretical foundations and data support for the design and application of semi-rigid base materials in asphalt pavements within arid desert regions.

Materials

Aeolian sand

The aeolian sand used in the experiment was taken from the Gurbantunggut Desert. The particle sizes of the aeolian sand were mainly concentrated between 0.075 mm and 0.3 mm. Aeolian sand was yellowish, as shown in Fig. 1(a). The inhomogeneity coefficient of the aeolian sand was 2.2. The curvature coefficient was 0.8, while the mass of particles larger than 0.075 mm accounted for 90.8% of the total mass. The scanning electron microscope diagram of the aeolian sand is shown in Fig. 1(b). The technical indexes are shown in Table 1.

Fig. 1

Aeolian sand.

Table 1 Aeolian sand technical indicators.

Cement

The ordinary Portland cement labeled as P.O42.5 made by Xinjiang Tianshan brand in Chinese is used, and the main technical indicators are shown in Table 2.

Table 2 Cement technical indicators.

Fly ash

The Class II F fly ash made by Xinjiang Longxin brand in Chinese is used. The main properties of the fly ash are shown in Table 3.

Table 3 Fly ash technical indicators.

Gravel

The technical specifications of the 10 mm-20 mm and 20 mm-30 mm coarse aggregates selected for this test are shown in Table 4.

Table 4 Gravel technical indicators.

Methodology

Overall procedure

To design the proportion of CFSAG mix with aeolian sand replaced by fine aggregate, gravel as coarse aggregate and cement fly ash as binding material. Firstly, the mechanical property change rules of CFSAG were studied under high-temperature 30℃, 40℃ and 50℃ maintenance. Secondly, freeze–thaw cycles tests in water and 2% Na₂SO₄ solution were executed 3rd, 7th, 11th and 15th to analyse the extent of the influence of aeolian sand mixing on the durability of CFSAG. The pH test inside the specimens under different working conditions was completed, and the strength evolution model was established. Finally, the surface damage characterisation study of specimens at 98% and 95% compaction degrees was carried out based on DIC, and the damage crack evolution pattern of the CFSAG specimen was derived. Figure 2 shows the technology roadmap of this work. Figure 3 shows the main test flow chart of this work.

Fig. 2

Technology roadmap.

Fig. 3

Overall flow chart.

Proportion of mixture

Firstly, based on pre-testing and previous research results³⁹. The fine aggregate is 100% replaced by aeolian sand. The cement and fly ash are used as the binding material, and a certain amount of gravel is calculated according to the internal addition method to complete the design of the proportion. Secondly, Based on the results of the previous studies³⁹ and field investigation of cement dosage for the base of highway roads, the cement dosage in this study is 4%. In order to further improve the utilisation of aeolian sand, the solid waste fly ash is selected for improvement. Fly ash dosage was designed according to the cement: fly ash = 1:3 based on pre-tests and reference to conclusions of the study by Lei⁴⁰, Liu³⁹, Liu⁴¹ and Chinese testing code JTG 3441⁴². According to the conclusion of Guo⁴³, the maximum dosage of aeolian sand in the base is 53%. To further improve the utilisation of aeolian sand and explore the effect of sand mixing on the road performance of mixes, the design used a maximum of 66% aeolian sand. It took a range of 8% of variations. According to material stabilised with inorganic binders with suspension structure with less than 50% gravel mixing, the dosage of gravel (10 mm ~ 20 mm: 20 mm ~ 30 mm = 1:1) was calculated to be 34%, 26% and 18% in order the internal addition method. Finally, 98% and 95% compaction degrees were used to conduct indoor experimental studies to verify that the CFSAG base meets the strength requirements of different road classes and traffic loads, the design results are shown in Table 5. Each proportioning mix code is designed based on the dosage of cement and the grade of aeolian sand (50%-A, 58%-B, 66%-C).

Table 5 Proportion of mixture design.

Proctor compaction test

The tests were carried out according to the Chinese testing code JTG 3441⁴⁴. The proctor compaction test was carried out according to the CFSAG base proportion designed in Table 5. Firstly, the amount of material used for cement, fly ash, aeolian sand and gravel at different moisture contents was calculated according to 98% and 95% compaction degrees. Secondly, five predetermined water contents (5%, 6%, 7%, 8% and 9%) are added to the calculated weighed dry material, then the mixture is placed in a black plastic bag for 4 h to finish infiltrating. Eventually, cement and fly ash were added to complete the proctor compaction test, and the compaction test results are shown in Fig. 4.

Fig. 4

The relationship between the maximum dry density and the optimum moisture content of each proportion of mixture.

High-temperature maintenance test

The Gurbantunggut desert is also known for its high temperature, with temperatures reaching extremes exceeding 50 °C. To more accurately simulate the actual temperature conditions during paving operations in this region, the maintenance temperature reference draws upon field survey data of construction site temperatures. Therefore, the effect of maintenance temperatures of 30 °C, 40 °C and 50 °C (hereafter referred to as high temperature) on the mechanical properties of CFSAG was carried out. The specimens with standard maintenance (temperature 20 ± 2 °C, humidity 95%) were used as a reference test group. According to the Chinese testing code JTG 3441⁴⁴, to complete the inner diameter of 150 mm × 150 mm cylindrical specimen production, the production of specimens is divided into two groups, each group of 6 specimens. One group was placed in a blowing drying oven at different temperatures to complete the high-temperature maintenance environment, and the other group was placed in standard maintenance to complete the reference. The day before the maintenance is for immersion maintenance to the age of 7d. Finally, the specimens that had reached the frequency of freeze–thaw cycles by an electro-hydraulic universal testing machine at a rate of 1 mm/min according to Chinese testing code JTG 3441⁴⁴.

Freeze–thaw cycle testing

The indoor freeze–thaw cycles tests were conducted in according to Chinese testing code GB/T 50082⁴⁵. Based on the regional temperature data and pavement monitoring results of the test section, the extreme minimum pavement temperature was found to be -18 °C, while the maximum temperature did not exceed 20 °C. Accordingly, the freeze–thaw temperature range was set at -18 °C to 20 °C. Each freeze–thaw cycle was set at 24 h, consisting of 16 h of freezing and 8 h of thawing. The temperature rise/fall rate within the chamber between -18 °C and 20 °C is approximately 8–10 °C per hour.

Following the aforementioned mix design, cylindrical specimens with an internal diameter of 150 mm × 150 mm were prepared according to the Chinese testing code JTG 3441⁴⁴. After 28 days of standard indoor curing, the specimens were divided into two groups. One group underwent freeze–thaw cycles (0, 3, 7, 11, and 15 cycles) in both clean water and a 2% Na₂SO₄ solution. The other group remained under standard curing as a control. Finally, the freeze-thawed specimens and the control group specimens were subjected to unconfined compressive strength tests at a loading rate of 1 mm/min.

pH test

The pH measurement of CFSAG mixes at the end of the freeze–thaw cycle was carried out using the water extraction method with reference to the research methods of related scholars^46,47,48. Firstly, the specimens that had reached the frequency of freeze–thaw cycles in water, 2% Na₂SO₄ environment were destroyed and used as a reference for the standard maintenance group of specimens. The destroyed specimen is divided into 3 equal parts. The height of the split is taken as the upper, middle and lower three positions (with 5 cm as the spacing), and the sample is selected from the middle position and evenly sampled. Secondly, the samples were taken and dried in an oven at 105 °C. After drying, the samples were ground using a mortar and pestle, and the powdered samples obtained after grinding were passed through a 0.3 mm sieve. The sieved powder was weighed to a mass of 10 g and then dissolved in 100 ml of distilled water with a stir, the upper suspension of the solution was taken after 2 min of stirring. Finally, sample pH values were determined using pH test paper and the mean value was calculated after repeating the test three times for each group of samples.

DIC test

The DIC equipment for this test was manufactured by Hefei Zhongke Junda Vision Technology Co. Firstly, the inner diameter 150 mm × 150 mm cylindrical specimen production and 7d standard maintenance (temperature 20 ± 2℃, humidity 95%) were completed according to the Chinese testing code JTG 3441⁴⁴. Secondly, the specimen that has reached the maintenance age will have black scattered spots on the surface and then be placed on the universal testing machine. The DIC equipment is placed in front of the specimen and calibrated, and the relevant parameters of the DIC equipment are regulated so that the equipment can accurately catch the black scattered spots. Finally, the DIC test and analysis system was initiated at the same time when the damage of compressive strength loading was started with 1 mm/min according to Chinese testing code JTG 3441³⁷. The strain and displacement on the surface of the specimen are measured at a frequency of 0.4 frame/second, and the compressive damage and displacement changes of the specimen are recorded.

Study of the road performance of CFSAG base

High-temperature maintenance test

Effect of maintenance temperature on compressive strength at different compaction degrees

High-temperature maintenance of CFSAG based on the compressive strength is shown in Fig. 5.

Fig. 5

Effect of temperature on compressive strength at different compaction degrees.

The compressive strengths of CFSAG mix 4-A, 4-B and 4-C showed a decreasing trend with increased aeolian sand dosage at different maintenance temperatures. The strength of CFSAG increases with increasing maintenance temperature and compaction degree when the aeolian sand substitution percentage is certain. High-temperature maintenance resulted in increased compressive strength of CFSAG mixes compared to standard maintenance conditions, and the 40 °C maintenance temperature was favourable for compressive strength formation. With the aeolian sand dosage of 50%, the compressive strength of CFSAG mix, which is 98% compacted degree, is 3.97 MPa, 4.26 MPa, 4.43 MPa, 4.38 MPa, respectively under standard, 30 °C, 40 °C and 50 °C maintenance. The compressive strength of CFSAG mix which is 95% compacted degree is 3.02 MPa, 3.21 MPa, 3.35 MPa, 3.31 MPa respectively under standard, 30 °C, 40 °C and 50 °C maintenance. When the compaction degree is 98%, the strength reduction relationship of CFSAG mixtures from 4-A to 4-C under different maintenance environments with increasing aeolian sand content is as follows: standard maintenance < 30 °C maintenance < 40 °C maintenance < 50 °C maintenance.

Under high-temperature maintenance conditions, compared with single-factor approaches using cement to enhance the compressive strength of aeolian sand mixtures¹⁹, the combined application of fly ash and coarse aggregates not only increases the compressive strength of the aeolian sand mixtures but also alters the mechanism by which high temperatures influence their compressive strength. This phenomenon arises from two factors: the formation of aggregates forming a skeletal structure within the mixture, and the pozzolanic effect of fly ash. High-temperature curing accelerates cement hydration, generating substantial Ca(OH)₂. The resulting alkaline environment promotes secondary reactions between fly ash and Ca(OH)₂, filling inter-aggregate voids and rendering the specimen’s internal structure more compact. According to the testing code JTG/T F20⁴², considering the high-temperature environment of summer deserts, when the cement content is 4% and the compaction degree is 98%, the admixture ratio of aeolian sand in CFSAG base courses can reach 58%.

The relationship between maintenance temperature and growth rate of compressive strength

In order to further quantify the extent to which high-temperature maintenance affects the compressive strength of CFSAG mixes, the compressive strength loss rate was defined, shown in Eq. (1).

$$Q_{r} = \frac{{\left( {C_{t} - C_{sc} } \right)}}{{C_{sc} }}$$

(1)

In Eq. (1):

$Q_{r}$: Compressive strength under high temperature maintenance (MPa).

$C_{sc}$: Compressive strength under standard maintenance (MPa).

The rate of loss on compressive strength in CFSAG mixes is shown in Fig. 6. The compressive strength increase between 20 °C ~ 40 °C was more significant than between 40 °C ~ 50 °C for CFSAG mixes. The growth rate of compressive strength was gradually decreased with the increase of aeolian sand dosage in CFSAG. When CFSAG specimens used 4% cement content and 98% degree of compaction, the compressive strength gain rates at 30 °C, 40 °C and 50 °C all exhibited the following sequence, Mix Proportion 4-A > 4-B > 4-C. When the compaction degree was reduced from 98 to 95%, the rate of loss of compressive strength under maintenance at 30 °C, 40 °C, and 50 °C increased gradually with the increase of aeolian sand dosage and increased with the decrease of the compaction degree. For CFSAG mix proportions 4-A, 4-B and 4-C. When maintained at 30 °C, the loss of unconfined compressive strength decreased by 24.65%, 28.49% and 31.88% respectively. When maintained at 40 °C, the loss of unconfined compressive strength decreased by 24.38%, 28.08% and 31.61% respectively. When maintained at 50 °C, the loss in ultimate compressive strength decreased by 24.43%, 28.38% and 31.80% respectively.

Fig. 6

The relationship between high temperature maintenance and the growth rate of compressive strength of the mixture.

High-temperature maintenance caused by CFSAG compressive strength growth rate of ‘first increase, then decrease’ rule of change and Zhou⁴⁹, Shi⁵⁰ study is consistent. The strength is increased mainly due to the volcanic ash activity effect and filling effect of fly ash. With the increase of the maintenance temperature, the fly ash activity degree is increased gradually, which stimulates the fly ash to undergo the alkali excitation stress and the density of the specimen structure is improved. Therefore, when maintaining specimens at 40 °C, the synergistic effect between accelerated cementation reactions due to high temperature and initial microstructural optimisation reached its peak, enabling the specimens to achieve optimum compressive strength at this temperature. There are two reasons for the further reduction in strength. Firstly, the high temperature accelerates the evaporation of free water inside the specimen structure, which reduces the free water required for the hydration reaction of cement and fly ash, resulting in a reduction of hydration products inside the specimen. Secondly, continued high temperatures lead to the evaporation of free water inside the specimen, making the structure produce capillary cracks and microporosity. The specimen gap in the process of testing mechanical properties due to vertical load will produce a tip stress concentration phenomenon, prompting the fine cracks and microporosity to carry out the specimen further so that the piece’s strength is decreased.

Freeze–thaw cycle test of CFSAG

Effect of the frequency of freeze–thaw cycles on compressive strength

The evolution of compressive strength in CFSAG specimens with different aeolian sand content over freeze–thaw cycles is shown in Figs. 7, 8, 9.

Fig. 7

Influence of freeze–thaw cycles on the compressive srength of 4-A with different compaction degrees.

Fig. 8

Influence of freeze–thaw cycles on the compressive strength of 4-B with different compaction degrees.

Fig. 9

Influence of freeze–thaw cycles on the compressive strength of 4-C with different compaction degrees.

The compressive strength of mix proportions 4-A, 4-B and 4-C were positively correlated with compaction degree with the same freezing and thawing cycles frequency at 4% cement dosage. The frost resistance of CFSAG gradually decreases due to the increase in the dosage of aeolian sand. The mix proportions 4-A, 4-B, and 4-C with freeze–thaw in water and 2% Na₂SO₄ compressive strength decrease gradually with the frequency of freezing and thawing. In contrast, the compressive strength under standard maintenance increases with age and the growth relationship is 4-A > 4-B > 4-C. At the end of the 3rd freeze–thaw cycle, the compressive strength difference between the CFSAG specimen at 98% compaction and the 4-A, 4-B, and 4-C specimens at 95% compaction was less than 0.1 MPa in both water and 2% Na₂SO₄ solution. At the 7th freeze–thaw cycle, the difference in compressive strength between the water environment and the 2% Na₂SO₄ environment significantly increased, with specimens in the 2% Na₂SO₄ environment exhibiting the most pronounced reduction in strength..

There are two main reasons for the slight variation in compressive strength exhibited by the CFSAG specimens following the 3rd freeze–thaw cycle in water and a 2% Na₂SO₄ solution. Firstly, the specimens exhibited minimal surface damage and appeared smooth after undergoing freeze–thaw cycles in water environment. The water environment had little effect of the overall compressive strength of the CFSAG specimens. Secondly, when specimens undergo salt corrosion and dissolution within the salt solution environment, SO₄²⁻ undergo secondary consumption reactions with the hydration reaction products within the specimen^51,52,53,54. This process fills the pores of the specimen, thereby enhancing its resistance to de-icing salt freezing. The specimens exhibit minimal variation in compressive strength at the macroscopic level.

The reason for the continuous decline in compressive strength of CFSAG specimens with increasing freeze–thaw cycles is as follows. As the content of aeolian sand gradually increases, it induces an ‘arch effect’⁵⁵, which increases the internal porosity of the specimens. This makes the specimens more susceptible to freeze–thaw action, leading to the aeolian sand and hydration products flaking off from the specimen surface. When the compaction degree is lower from 98 to 95% and 15th freeze–thaw cycles are completed for CFSAG. The compressive strength of mix proportions 4-A, 4-B and 4-C in standard maintenance were reduced by 0.6 MPa, 0.5 MPa, and 0.4 MPa. The compressive strength of mix proportions 4-A, 4-B and 4-C in water freeze–thaw environment were decreased by 0.5 MPa, 0.5 MPa, and 0.4 MPa. The compressive strength of proportion 4-A, 4-B and 4-C in 2% Na₂SO₄ solution freeze–thaw environment were reduced by 0.5 MPa, 0.4 MPa, and 0.4 MPa.

Effect of the frequency of freeze–thaw cycles on the quality change in specimens

The evolution of the quality change in the CFSAG specimens in water after freeze–thaw cycles is shown in Fig. 10.

Fig. 10

Effect of water freeze–thaw cycles on the proportion quality at different compaction degrees.

After undergoing the third freeze–thaw cycle in a water environment.The quality growth rates of specimens with mix proportions 4-A-98%, 4-B-98% and 4-C-8% were 5.9%, 5.1% and 4.5% respectively. Specimens with mix proportions 4-A-95%, 4-B-95% and 4-C-95% exhibited quality growth rates of 4.9%, 4.2% and 3.8% respectively. When undergoing 15th freeze–thaw cycles, the quality loss rates of specimens with mix proportions of 4-A-98%, 4-B-98% and 4-C-98% were 4.5%, 5.4% and 6.2% respectively. Specimens with mix proportions of 4-A-95%, 4-B-95% and 4-C-95% exhibited quality loss rates of 6.7%, 7.4% and 8.8% respectively. During the freeze–thaw cycle, the quality of each mix proportion specimen exhibited a trend of initial increase followed by decrease. This phenomenon is primarily attributable to two factors: firstly, the secondary hydration reaction between fly ash and the hydration product Ca(OH)₂, whose reaction products caused an increase in specimen quality at the initial stage. On the other hand, the initial freezing and thawing of the specimen surface is very smooth. With the increasing times of freezing and thawing, the aeolian sand starts to flake off significantly, resulting in the generation of hydration products less than the amount of aeolian sand flaking off the specimen surface.

The evolution of the quality change in the CFSAG specimens in 2% Na₂SO₄ solution after freeze–thaw cycles is shown in Fig. 11. The quality changes at the 7th freeze–thaw cycle for the 98% compaction degree proportions 4-A, 4-B and 4-C were less than the 95% compaction degree. The leading cause of the above phenomenon is the compaction degree of 98% of the specimen, and the overall compaction performance is higher than 95%, resulting in SO4^2- through the pore into the specimen inside the amount of less than 95%. At the same time, SO4^2- reacted with the hydration products inside the specimen to generate expansive material, which destroys the bond between the aeolian sand and coarse aggregate. This causes the aeolian sand and binder paste on the specimen’s surface to progressively spall during freeze–thaw cycles, resulting in a continuous increase in quality loss. After the 15th freeze–thaw cycle in 2% Na₂SO₄ environment, compared with the water environment. The same feature is that the changing trend of specimen quality is the same, both show the trend of ‘first increase and then decrease’. The difference is that the rate of quality loss of the specimens in 2% Na₂SO₄ solution is lower than that of the water environment in the early phase of the freeze–thaw cycle. In contrast, the quality loss gradually decreases in the final phase due to the sulphate erosion effect.

Fig. 11

Effect of Na₂SO₄ freeze–thaw cycles on the proportion quality at different compaction degrees.

To further investigate the patterns of appearance changes in CFSAG proportion mix specimens under different freeze–thaw environments, and considering page limitations and the consistency in appearance trends across all proportion mix specimens. This study selected the 4-B proportion mix specimen for representative analysis.

Figure 12 shows the appearance changes of the specimen after undergoing different numbers of freeze–thaw cycles in water. By the 3rd and 7th freeze–thaw cycles, surface mortar spalling was not yet pronounced, with no significant pitting or cracking observed. Following the 11th cycle, the binder slurry in the upper-middle section of the specimen began to gradually detach. After the 15th freeze–thaw cycle, the surface slurry had largely detached, the aeolian sand begins to flake away, revealing pitted grooves on the specimen’s surface, though the aggregate remained unexposed.

Fig. 12

Effect of the number of freeze–thaw of 4-B water on the appearance of specimens. The surface defects on the specimen: Slurry spalling, Smooth, pitted grooves.

Figure 13 depicts the appearance changes of specimens after freeze–thaw cycles in a 2% Na₂SO₄ solution. After the 3rd freeze–thaw cycle, significant spalling appeared on the specimen surface, with extensive loss of binder slurry in the upper-middle section. By the 7th freeze–thaw cycle, surface slurry spalling intensified further, exposing coarse aggregates, though the specimen retained overall integrity without visible cracks. After the 11th cycle, localised pitting and salt out appeared on the surface, though the specimen retained its overall shape. Following the 15th freeze–thaw cycle, the surface slurry had almost entirely detached, the freeze–thaw layer began to peel away, and the aggregate became distinctly exposed.

Fig. 13

Effect of the number of freeze–thaw of 4-B 2% Na₂SO₄ solution on the appearance of specimens. The surface defects on the specimen: Smooth, Slurry spalling, Pitted grooves, Exposed aggregate, Salt out.

Compared to the aqueous environment, the spalling of surface slurry and aeolian sand was more pronounced in the 2% Na₂SO₄ solution. Under the same freeze–thaw cycles, specimens exhibited more severe visual damage and significantly greater coarse aggregate exposure than in the aqueous medium. Visual crack morphology revealed damage primarily manifested as triangular or vertical strip-like fissures, consistent with DIC testing conclusions.

Evolution model of freeze–thaw cycles frequency and compressive strength of aeolian sand

In order to evaluate the freeze–thaw resistance of CFSAG specimens in water and 2% Na₂SO₄ solution. Through calculating the compressive strength coefficient of specimens under different freeze–thaw environments, the variation patterns in the mechanical properties of CFSAG specimens are quantitatively characterised.

The compressive strength coefficients for different freeze–thaw environments were calculated according to Eq. (2) and Eq. (3).

The water compressive strength coefficient:

$$P_{w} = \frac{{K_{w} \left( N \right)}}{{K_{0} }}$$

(2)

Formula, $P_{w}$ is water compressive strength coefficient, %, $K_{w}$ is water erosion compressive strength, MPa, $K_{0}$ is standard maintenance compressive strength, MPa.

The 2% Na₂SO₄ compressive strength coefficient:

$$P_{s} = \frac{{K_{s} \left( N \right)}}{{K_{0} }}$$

(3)

Formula, $P_{s}$ is 2% Na₂SO₄ compressive strength coefficient,%, $K_{s}$ is 2% Na₂SO₄ erosion compressive strength, MPa, $K_{0}$ is standard maintenance compressive strength, MPa.

The relative compressive strength coefficients of CFSAG specimens under the freeze–thaw cycle can be calculated according to Eq. (2) and Eq. (3).

$P_{w}$, $P_{s}$ show a better quadratic polynomial relationship with the frequency of freeze–thaw cycle erosion N. The quadratic polynomial expression is proposed as the relationship as shown in Eq. (4) and Eq. (5).

$$P_{w} = aN^{2} + bN + c$$

(4)

$$P_{s} = aN^{2} + bN + c$$

(5)

Formula, a, b and c are empirical constants. As a result of the freeze–thaw cycle erosion cycle number N = 0, $P_{w}$ or $P_{s}$ = 1, thus c = 1.

The relative compressive strength coefficients of the specimens according to Eq. (4) and Eq. (5) and the fitting equations of the specimens are obtained as shown in Table 6. The aeolian sand dosage parameter $\omega$, combined with the coefficients of compressive strength $P_{w}$ , $P_{s}$ and the frequency of freezing and thawing N to conduct regression analyses. The results are shown in Figs. 14 and 15.

Table 6 Regression analysis of freeze–thaw cycles.

Fig. 14

Relationship between the number of freeze–thaw cycles in water, $\omega$, and $P_{w}$.

Fig. 15

Relationship between the number of freeze–thaw cycles in Na₂SO₄, $\omega$, and $P_{w}$.

Further consideration is needed to determine the relatedness between aeolian sand dosing and relative compressive strength coefficients. The quadratic polynomial functions were established by using the aeolian sand dosage as the independent variable and the coefficients of the fitted equations as the dependent variables, and were respectively substituted into the corresponding fitted functions.

The evolution model Eqs. (6)-(9) can be obtained as relative compressive strength coefficients evolution of CFSAG under different compaction degrees in water and Na₂SO₄ freeze–thaw environments.

Water freezing and thawing with 98% compaction degree:

$$P_{w} = \left( { - 0.0016\omega^{2} + 0.0023\omega - 0.0011} \right)N^{2} + \left( {0.0024\omega^{2} - 0.00189\omega + 0.0058} \right)N + 1$$

(6)

Water freezing and thawing with 98% compaction degree:

$$P_{w} = \left( {0.0116\omega^{2} - 0.0122\omega + 0.0030} \right)N^{2} + \left( { - 0.1184\omega^{2} + 0.1058\omega - 0.0288} \right)N + 1$$

(7)

2% Na₂SO₄ freezing and thawing with 98% compaction degree:

$$P_{w} = \left( { - 0.0084\omega^{2} + 0.0108\omega - 0.0037} \right)N^{2} + \left( {0.1470\omega^{2} - 0.1943\omega + 0.0560} \right)N + 1$$

(8)

2% Na₂SO₄ freezing and thawing with 95% compaction degree:

$$P_{w} = \left( { - 0.0057\omega^{2} + 0.0072\omega - 0.0024} \right)N^{2} + \left( {0.0992\omega^{2} - 0.1367\omega + 0.0358} \right)N + 1$$

(9)

The results of the fitted models in water and 2% Na₂SO₄ environments were compared with the test values for the different compaction coefficients in Figs. 16, 17, 18, 19. The compressive strength coefficients of CFSAG under freeze–thaw cycling in water and 2% Na₂SO₄ solution are closely distributed on the surface of the 3D model. The chi-square test of the fitted model is very low, which indicates that the accuracy of the fitted model is high. The research model provides a foundation for predicting engineering application metrics within this material system. The modeling-related parameters are shown in Table 7.

Fig. 16

Relationship between the frequency of freeze–thaw cycles,$\omega$, and $P_{w}$ for 98% water.

Fig. 17

Relationship between the frequency water freeze–thaw cycles, $\omega$, and $P_{w}$ for 95% water.

Fig. 18

Relationship between the frequency of freeze–thaw cycles,$\omega$, and $P_{w}$ for 98% Na₂SO_4.

Fig. 19

Relationship between the frequency of freeze–thaw cycles,$\omega$, and $P_{w}$ for 95% Na₂SO_4.

Table 7 Goodness of fit.

Effect of freezing and thawing environments on pH in specimens

The evolution of pH in the CFSAG specimens under freeze–thaw cycles of water, 2% Na₂SO₄ and the reference group of standard maintenance is shown in Figs. 20, 21, 22.

Fig. 20

Effect of standard maintenance environment on the pH of the mix proportion at different compaction degrees.

Fig. 21

Effect of freezing and thawing of water on the pH of the mix proportion at different compaction degrees.

Fig. 22

Effect of Na₂SO₄ freeze–thaw on the pH of the mix proportion at different compaction degrees.

The pH values in water freeze–thaw environments tended to decrease more slowly than in the standard maintenance specimens at the same age. The difference in pH within the identical proportion test specimens was close to 3.5. The trend of pH value changes under different compaction degrees is consistent, and the pH value variation is more evident due to the dosage of aeolian sand. The pH value variation trend of CFSAG mixes under two compaction degrees shows 4-A > 4-B > 4-C. The pH values in 2% Na₂SO₄ freeze–thaw environments showed a slightly smaller tendency to decrease compared to the standard maintenance specimens at the same age.

With increasing freeze–thaw cycles, there were two reasons which caused the pH value of the specimens in the Na₂SO₄ environment to be slightly higher than the pH value in the standard environment. First, the freeze–thaw cycle makes the surface appearance of the specimen smooth and gradually rough. The Na⁺ and SO4^2- in the solution during the water solubility process penetrate the structure through the surface pores of the specimen and react with the hydration product Ca(OH)₂. So that the amount of OH^- generated inside the specimen increases. Secondly, the volcanic ash reaction of fly ash consumes OH^- cations, which precipitate from the reaction of the hydration products Ca(OH)₂ and Na₂SO₄. This ultimately resulted in a slightly lower reduction of OH^- cation content inside the specimen than in the standard maintenance. In general, the pH value of the specimens gradually decreased with the increase in the maintenance age and the frequency of freeze–thaw cycles. The main reason is that the decomposition rate of C-S–H hydration products forming new Ca(OH)₂ under freeze–thaw cycling conditions is lower than the dissolution rate of Ca(OH)₂.

Study on CFSAG crack extension pattern based on DIC technique

Strain pattern evolution characteristics with different proportions

The strain patterns of CFSAG specimens at the different phases of the compressive strength damage process are shown in Figs. 23, 24, 25, 26, 27, 28. According to the changing results of the colour gradient of the strain pattern of the CFSAG specimen during the test. The development process of the surface cracks of the specimen under 98% and 95% compaction degrees were divided into stage Ⅰ (initial phase), stage Ⅱ (developmental phase), and stage Ⅲ (extension-run through phase).

Fig. 23

98%-4-A proportion compressive strength damage process strain clouds.

Fig. 24

98%-4-B proportion compressive strength damage process strain clouds.

Fig. 25

98%-4-C proportion compressive strength damage process strain clouds.

Fig. 26

95%-4-A proportion compressive strength damage process strain clouds.

Fig. 27

95%-4-B proportion compressive strength damage process strain clouds.

Fig. 28

95%-4-C proportion compressive strength damage process strain clouds.

As shown in the initial stage of Figs. 23, 24, 25, 26, 27, 28, the colour gradient distribution of the strain patterns of specimens of proportions 4-A, 4-B and 4-C are consistent. The specimens show the phenomenon of longitudinal extension along both sides of the surface, and the strain focus at the centre is not apparent. This indicates at this stage that the degree of compaction and the proportion of aeolian sand incorporated have little effect on specimen deformation. There was also no crack on the surface of the specimen, and the initial strain distribution of the specimen at 98% compaction degree was in the range of 1100 ~ 1500 με, and that of the specimen at 95% compaction degree was in the range of 1200 ~ 1600 με. As shown in the developmental phase of Figs. 23, 24, 25, 26, 27, 28, the strain patterns of the specimens of proportion 4-A, 4-B and 4-C all showed strain concentration bands, and the strain colour gradient variations exhibited differently for each proportion. The region of strain concentrating on the surface of the specimen extends from the upper side to the lower side, which indicates a gradual increase in the internal deformation of the mixture under sustained loading. The length of the strain concentration bands for each mix proportions increases gradually with the increase of the aeolian sand dosage. There is a more significant effect of the surface crack evolution characteristics due to the decrease of the compaction degree. The strain range for crack initiation at 98% compaction degrees is 3500 με to 5500 με, and for 95% compaction degrees, it is 4400 με to 6700 με. As shown in the extension-run through phase of Figs. 23, 24, 25, 26, 27, 28, the specimen reaches the peak loading process, and the discrete partial strain concentration areas lap and merge with each other along the loading direction, forming a high strain concentration band. Micro-cracks from short cracks converge into long cracks, the red stripe region continues to expand the macro-expression of crack widening. Then, macro destruction of cracks along the high strain concentration band rapidly cracks and runs through the specimen. After a brittle fracture, the specimen loses the ability to withstand compressive stress. At the end of specimen destruction, the micro-strains of proportions 4-A, 4-B, and 4-C at 98% compaction degrees were 11,063.8 με, 12,191.2 με, and 13,885.5 με. At the end of specimen destruction, the micro-strains of proportions 4-A, 4-B, and 4-C at 95% compaction degrees were 13,754.9 με, 15,351.0 με, and 16,806.2 με. This shows that the strain at the end of the destruction of the specimen is inversely proportional to the compressive strength.

Comparison of the strain evolution patterns of surface cracks in CFSAG specimens at 98% and 95% compaction degrees. The similarity is that specimens under different compaction degrees in the compressive strength damage process of crack expansion are from a single main crack that began to extend. The concentration and development of stress concentration regions on the specimens’ surfaces intensified with increasing aeolian sand content. Specimens of different mix proportions exhibited similar damage patterns under the same compaction degree. The difference is that the strain concentration band of the specimen at 95% compaction degree develops faster. The strains on the surface of the specimen appear to cross each other. In contrast, the strain on the surface of the specimen at 98% compaction degree shows a single longitudinal crack and begins to expand until it runs through the whole specimen. There are two reasons for the differences in strain development in the specimens. The first one is because CFSAG mixtures are a non-homogeneous material with anisotropic, non-linear and variable characteristics. The continuous loading resulted in an uneven distribution of vertical strain in the specimens with different aeolian sand dosages. Second, the specimen’s internal aggregate and pore distribution degree differ under different aeolian sand dosages. After the initial crack formation, the specimen with a smaller dosage of aeolian sand requires more energy during the microcrack development period before destruction. Macro-scale behaviour exhibits enhanced load-bearing capacity without rapid complete rupture. The specimen with a larger dosage of aeolian sand mixture microcracks form and develop into macroscopic cracks very quickly after the formation of microcracks. In test specimens composed of mixtures with high aeolian sand content, microcracks rapidly propagate into macroscopic cracks once formed. Crack initiation originates from gravel particles with diameters of 10–20 mm and 20–30 mm at the specimen’s core, as microcracks cannot penetrate these hard particles during propagation. Subsequently, the mixture extends further towards the edges of the hard particles within the relatively low-strength microcrack zones, ultimately rupturing along the binder-aggregate interface.

Comparing the specimen photographs at the end of the breakage, as shown in the destruction of photographs of Figs. 23, 24, 25, 26, 27, 28. The shape and location of the main crack of the specimen coincide with the strain concentration region in the strain distribution cloud figure (red or missing area in the cloud figure). In the compressive strength damage test, the randomness of the cracking point of the specimen is large, and the aggregation of early microcracks causes the formation of macroscopic cracks. The formation of microcracks is due to the specimen in the loading process. When the load is greater than the aggregate bearing ability, the cement-fly ash binding slurry around the coarse aggregate begins to slide. It was found that the compressive strength test of the CFSAG specimen of proportions 4-A, 4-B, and 4-C is subjected to balanced force, the specimen is not eccentrically subjected to compression, and the value of the compressive strength force is accurate. Finally, the distribution of aggregate and bonding slurry in the region of strain concentration and the region of crack distribution after the specimen’s damage will be compared. It was concluded that the shape of the main crack in the compressive test process of CFSAG specimens with different segregation levels was related to the aeolian sand and the binding material slurry in the force concentration region. Thus, the order of magnitude of micro-strains affecting CFSAG mixes is aeolian sand > binding material slurry > gravel.

Analysis of crack extension in CFSAG

The transverse and vertical displacements of the points selected on the intercept line in the main crack development path of the specimen which is the region of DIC calculation are shown in Figs. 29, 30.

Fig.29

Transverse displacement.

Fig.30

Vertical displacement.

The two points adjacent to the middle of the intercept line (near the horizontal coordinate of 5 mm) generate a large difference between the vertical and transverse displacements. The difference in the displacements increases with the loading process, indicating that the main crack of the damaged specimen is located here. The relative displacement of points on either side of the crack in the transverse direction is the transverse displacement, and the relative displacement in the vertical direction is the vertical displacement. It can be concluded that the CFSAG specimens generate both transverse and vertical displacement during the compressive damage process.

At the initial stages with compaction degrees of 98% and 95% respectively. During the compressive breakage process of specimens from mix proportions 4-A, 4-B, and 4-C, both the transverse and vertical displacement variations at each monitoring point along the crack intercept line were close to zero, with no apparent jumping phenomenon as shown in Figs. 29, 30. The surface points were moved as a whole due to the vertical load, and no cracks were formed in the specimens. During the development phase, the transverse and vertical displacements at adjacent points in the middle of the intercept line generate a significant difference. As the loading continued, the surface of the specimen began to shift with the aeolian sand and the bonding material slurry, and cracks began to form.

The vertical displacements show an overall increase, with insignificant changes in the displacements at both ends of the intercepts. From Fig. 29, it can be seen that the dosage of aeolian sand significantly affects the formation of cracks in the specimens. From Fig. 30, it can be seen that the reduction of compaction degree also eases the formation of cracks. In the extension-run through phase, the difference between the transverse and vertical displacements of the adjacent points at the intercept centre position gradually increases. This indicates that the overall bonding of the surface layer of the specimen begins to decrease, and cracks on the surface of the specimen begin to form massively. As the load increases, the microcracks on the surface of the specimen begin to connect. In the extension-run through phase, microcracks continue to extend and expand. The transverse and vertical displacement curves on the ‘step’ span height have reached the maximum, and the specimen surface cracks have been formed. Finally, the specimen under the impact of the vertical load is directly damaged.

At 98% compaction degree, the transverse displacements of the specimen of proportion 4-A at two points near the surface cracks during the extension-run through phase were respectively 0.27241 mm and 0.15682 mm, and the vertical displacements were respectively 0.07762 mm and 0.03393 mm, as shown in Figs. 29 (a), 30 (a). The transverse displacements of the specimen of proportion 4-B at two points near the surface cracks during the extension-run through phase were respectively 0.34049 mm and 0.29009 mm, and the vertical displacements were respectively 0.07753 mm and 0.00079 mm, as shown in Figs. 29 (b), 30 (b). The transverse displacements of the specimen of proportion 4-C at two points near the surface cracks during the extension-run through phase were respectively 0.06137 mm and 0.10971 mm, and the vertical displacements were respectively 0.47316 mm and 0.16061 mm, as shown in Figs. 29 (c), 30 (c).

At 95% compaction degree, The transverse displacements of the specimen of proportion 4-A at two points near the surface cracks during the extension-run through phase were respectively 0.02963 mm and 0.03976 mm, and the vertical displacements were respectively 0.08189 mm and 0.00444 mm, as shown in Figs. 29 (d), 30 (d). The transverse displacements of the specimen of proportion 4-B at two points near the surface cracks during the extension-run through phase were respectively 0.08501 mm and 0.03295 mm, and the vertical displacements were respectively 0.63016 mm and 0.05829 mm, as shown in Figs. 29 (e), 30 (e). The transverse displacements of the specimen of proportion 4-C at two points near the surface cracks during the extension-run through phase were respectively 0.24322 mm and 0.06396 mm, and the vertical displacements were respectively 0.31896 mm and 0.03224 mm, as shown in Figs. 29 (f), 30 (f).

In summary, the comparative analysis of the strain–displacement changes of CFSAG specimens under different loads can be classified into elastic, plastic and damage stages based on the evolution of the displacement increment curves. The displacement increment curves at both ends of the elastic phase intercept show an approximate positive correlation with the load, and the displacement field exhibits an elastic response. In the plastic phase, the transverse and vertical displacement increment curves appear as ‘steps’, and the displacement field is gradually transformed from linear to non-linear. In the damage stage, when the peak load is reached, the displacement increment curve of the specimen is singularly discontinuous, and the macroscopic damage crack is extended from the abscissa of 0 mm to both ends until the specimen is damaged.

Feasibility of aeolian sand replacing fine aggregates in asphalt pavement base

Investigation of roads in circum-desert regions has revealed widespread occurrences of pavement heaving. This study identifies the freeze–thaw cycles of saline solutions and temperature fluctuations as the primary causes of this distress. To address this issue, we conducted a comprehensive study on the durability and specimen deformation of a cement-fly ash stabilized aeolian sand gravel base with a 4% cement dosage. The compressive strengths for various mix proportions with different aeolian sand contents were determined at 98% and 95% compaction degrees, providing a basis for selecting appropriate designs for different road classes in desert environments. All experiments were performed in accordance with the Chinese test specification JTG/T F20⁴².

Beyond meeting fundamental mechanical performance criteria, the study further evaluated the durability of the material under both high-temperature and freeze–thaw environments. Given that the maximum temperature during the curing period for semi-rigid bases in desert areas can reach 70℃—a condition noted in existing research^56,57 to adversely affect the strength of aeolian sand mixtures—this study specifically investigated the variation pattern and feasibility of base course strength under different curing conditions. To counteract the potential strength loss, fly ash and coarse aggregates were incorporated into the mix. Our results demonstrate that these additions significantly enhance the compressive strength, a finding which is consistent with previous research^58,59,60.

Secondly, freeze–thaw cycle tests were conducted to investigate the pavement heaving distress induced by sulfate environments. The results revealed that sulfate erosion has a relatively minor impact on the strength of the cement-fly ash stabilized aeolian sand-gravel mixture during the initial stages. After undergoing freeze–thaw cycles, the mixture’s strength could still reach levels comparable to that of regenerated specimens of the same age under standard curing, which aligns with the findings reported in the literature^61,62,63,64.

Consequently, based on fulfilling the requirements for mechanical performance and durability, and adhering to the principle of high aeolian sand content, low binder dosage, and superior economic efficiency, the cement-fly ash stabilized aeolian sand-gravel mixture was selected according to Table 8 for use in base and sub-base courses of different traffic levels and main arteries. For Expressways and first-class highways with medium-to-light traffic volumes, 66% aeolian sand content is recommended for the semi-rigid base course of asphalt pavements. This approach increases the utilization of solid waste and aeolian sand, thereby contributing to the mitigation of global warming.

Table. 8 The 7 d unconfined compressive strength of cement-and fly ash-stabilized materials (MPa).

Discussion

This paper focuses on the road performance of 100% replacement of fine aggregate aeolian sand with semi-rigid base layers for asphalt pavements. The durability properties of CFSAG have been investigated using high-temperature maintenance and freeze–thaw cycle tests. The DIC test analysed the causes and factors affecting the crack extension of subgrade by the change of aeolian sand dosage under different compaction degrees. It is considered feasible for the CFSAG base. This study concluded that aeolian sand has a significant negative impact on durability performance and should be reasonably incorporated according to the design requirements, which is consistent with the results of previous studies^65,66. The appropriate increase in maintenance temperature is favourable to the formation of strength in CFSAG specimens, which agrees with the results of previous studies^25,67,68. Fly ash improves the sulphate erosion resistance of CFSAG mixes, which is consistent with the results of other earlier studies^69,70,71,72.

Compared with previous related studies^73,74,75, the compressive strength of this test after 7d of high-temperature maintenance was more than 2.8 MPa with standard maintenance at a maximum aeolian sand dosage of 66% and cement content of 4%, the results of the study meet the requirements of the design and acceptance indexes for low-traffic-volume base structures of highways⁴², which fully proves the results of the study apply to semi-rigid base structures of asphalt pavements in the desert hinterland. The main reason why the strength of aeolian sand mixtures increases under high-temperature maintenance is that the fly ash activity degree gradually increases, which stimulates the alkali excitation reaction of fly ash and improves the densification of the specimen structure, which is in agreement with the research results^76,77.

The main reason for the more minor damage of the specimen after the 15th freeze–thaw cycle is that the larger dosage of aeolian sand and fly ash, the smoother appearance of the specimen and the lesser amount of internal porosity, which slowed down the erosion of the specimen. It was found that the frost resistance of CFSAG is better than that of the conventional semi-rigid base^41,78,79,80. After the end of the destruction, the internal pH of the specimen gradually decreased, and the specimen showed the trend of ‘increasing first and then decreasing’ in both water and 2% Na₂SO₄, which was consistent with the research results⁴⁶. Based on the freeze–thaw cycle test study, the CFCSAG freeze–thaw resistance evolution model was established by considering the parameters affecting the design and construction (weathered sand, compaction and compressive strength), which provides the data calculation method for the subsequent related research. Based on DIC technology, specimens’ compressive strength damage process under different aeolian sand proportions and compaction degrees is monitored, which studies the crack formation and damage mechanism of aeolian sand mixtures from the macro displacement and strain perspectives. The study’s results reveal the destruction mechanism of the macroscopic mechanical properties of CFSAG specimens. The strain evolution pattern and the change rule of transverse and vertical displacements in the damage process of the specimen are similar to the results of related studies^{81,82,83,84,85} .

Based on the limitations of current research results, investigations may be conducted into the trend of strength changes under extreme high-temperature maintenance conditions. The coupled effect of freeze–thaw cycle and high-temperature maintenance on semi-rigid base courses of asphalt pavements should be studied. At the same time, the theoretical study of multi-factor coupling should be conducted to establish more applicable predictive models. This will optimise wind-blown sand and identify low-carbon, environmentally sustainable, and economical road construction materials for arid desert regions. It is recommended that test roads with different compaction degrees and aeolian sand dosages be paved in arid desert areas, and long-term monitoring should be carried out to explore the effectiveness of the engineering application of cemented fly ash-stabilised aeolian sand gravel base.

Conclusions

To investigate the durability performance and crack extension pattern of CFSAG base. The effects of high summer temperatures and winter freeze–thaw action on semi-rigid base layers of asphalt pavements in desert regions were simulated by indoor tests. High-temperature maintenance tests of CFSAG at 30 °C, 40 °C and 50 °C were carried out. The freeze–thaw resistance of CFSAG was analysed in fresh water, and 2% sodium sulphate environments at temperatures between -20 °C ~ 20 °C, and an evolution model for freeze–thaw compressive strength was established combining aeolian sand dosage. The crack extension pattern and deformation characteristics of CFSAG specimens with different aeolian sand dosages and compaction degrees were clarified. The main conclusions obtained are as follows.

• 1) Compared with the standard maintenance conditions, high-temperature maintenance improved the compressive strength of CFSAG, and the 40 °C maintenance conditions were most favourable to the formation of the pre-strength of CFSAG. At the same cement dosage, when the maintenance temperature is the same, the strength of the CFSAG mix decreases gradually with the increase of the aeolian sand dosage. The strength of CFSAG mixes increases with rising maintenance temperature and increasing compaction degree when the amount of aeolian sand is certain.

• 2) The compressive strength of CFSAG mixes decreases with an increasing number of freezes and thaws during freeze–thaw cycles in water and 2% Na₂SO₄ solution environment. After the 7th freeze–thaw cycle, salt corrosion and freeze–thaw causes resulted in a gradual increase in the difference in compressive strength of the specimens in water and 2% Na₂SO₄ with the increase in the frequency of freeze–thaw cycles.

• 3) By testing the specimen quality and internal pH after different numbers of freeze–thaw cycles, It was obtained that with the increase of the dosage of aeolian sand and the decrease of compaction degree, the spalling degree of aeolian sand and bonding material slurry on the surface of the specimen was significantly increased by the frequency of freezing and thawing cycles and the influence of the environment. The specimen quality increment increased and then decreased with the frequency of freezing and thawing. The specimen in the water freeze–thaw, 2% Na₂SO₄freeze-thaw and standard maintenance of the same age under the pH reduction trend relationship: Standard maintenance > Water > 2% Na₂SO₄, mainly because the Na⁺, SO4^2- through the specimen surface pores penetrate the structure of the internal, and the hydration product of Ca(OH)₂ reaction occurs to increase the specimen internal generation of OH^- generation.

• 4) To establish a compressive strength evolution model for CFSAG using the frequency of freezing and thawing cycles, compressive strength coefficient, and aeolian sand dosage of CFSAG specimens. The ReducedChi-Sqr and SSR of the test values and the calculated values of the simulated model are close to 1, R² minimum 0.9885, which indicates that the model can better describe the strength evolution rule of CFSAG mixes under different compaction degrees and different freeze–thaw cycle environment.

• 5) Cement-fly ash co-reinforcement had a specific effect on delaying the cracking and expansion of the aeolian sand mixture. The cracking and destabilisation toughness of cement-fly ash co-reinforcement were positively correlated with the compaction degree and gravel dosage and negatively correlated with the dosage of aeolian sand. During the crack extension process, the number of crack strips on the specimen surface and the total length of crack development overall tended to expand with the increase in the dosage of aeolian sand. When the CFSAG proportion is the same, and the compaction degree is different, the crack width and extension-run through increase with decreasing compaction degree.