Long-term skid and abrasion resistance of limestone aggregate asphalt mixtures

1 Introduction

The skid resistance and abrasion resistance of asphalt pavement are critical indicators to ensure driving safety, which directly affect the braking performance and driving stability of vehicles (Edwin et al., 2022; Xie et al., 2024). In the construction of expressway asphalt pavements, hard rocks such as basalt and diabase are usually used as aggregates for the surface layer. Such aggregates have high mechanical strength and excellent abrasion resistance, which can well meet the requirements of skid resistance and durability for the pavement surface layer (Chen et al., 2025; Meng et al., 2024). However, hard rock resources are unevenly distributed in China: limestone resources are abundant in some regions, while high-quality hard aggregates such as basalt are relatively scarce. If the traditional material selection method is still adopted to transport aggregates from other regions over long distances, it will not only significantly increase the project cost, but also bring higher energy consumption and carbon emissions, resulting in a considerable economic and environmental burden. Therefore, jointly driven by regional aggregate resource constraints and the imperatives of green transportation development, substituting conventional hard aggregates with local limestone aggregate in the surface wearing course of asphalt pavements not only offers significant economic and environmental benefits, but also represents an urgent task that must be advanced promptly.

Limestone is a sedimentary rock material with abundant reserves and wide distribution, which has the advantages of low cost, easy processing, and good adhesion to asphalt. Nevertheless, its crushing value, abrasion value, and polishing resistance are generally inferior to those of hard aggregates such as basalt (Meng et al., 2024; Wang et al., 2026; Zong et al., 2022). In recent years, numerous studies have been carried out by scholars on the application of limestone asphalt mixtures. (Yu et al., 2025) investigated the evolution of asphalt surface texture and variations in skid and wear resistance using a self-developed laboratory accelerated abrasion apparatus. (Ran et al., 2025) adopted three-dimensional reconstruction techniques to examine the changes in skid and wear resistance of thin-layer asphalt mixtures after wheel-induced abrasion. With regard to the mesoscopic deformation and morphological characterization of asphalt materials, (Zhu et al., 2026) utilized digital image correlation (DIC) to precisely measure the deformation characteristics of asphalt mixtures under different compaction conditions, thus offering a feasible technical pathway for the quantitative analysis of surface morphological evolution in asphalt pavements. (Yu et al., 2023a) found that the bonding capacity between asphalt and limestone is 2–3 times that with granite, suggesting that alkaline aggregates such as limestone should be included in raw materials for pavement engineering. (Zhang et al., 2025) further explored the influence of partially replacing limestone aggregate with steel slag on the interfacial interaction between asphalt and aggregate. The results indicated that regulating aggregate composition can effectively optimize the interfacial adhesion performance of limestone-based asphalt systems. Since interfacial adhesion directly determines the resistance of aggregate to stripping under repeated wheel abrasion, this study provides a reference for improving the long-term skid and wear resistance of limestone asphalt pavements. (Xu et al., 2021) explored the application of limestone in the bottom layer of double-layer permeable asphalt pavement. (Zhang et al., 2022) investigated methods to improve the abrasion resistance of limestone aggregate and demonstrated that the use of lithium silicate mixed solution (LSMS) and lithium silicate mixed solution (LALS) can enhance the abrasion resistance of limestone aggregate. (Qi et al., 2023) employed an accelerated abrasion machine to simulate the abrasion process of asphalt pavements and found that limestone exhibited the poorest skid resistance and stability among all tested aggregates. Existing studies have mainly focused on the skid resistance, pavement performance, and mechanical properties of limestone asphalt pavements, laying a foundation for the application of limestone in asphalt pavements. However, several shortcomings still exist. On the one hand, there is a lack of research on the long-term abrasion resistance and its evolution mechanism of limestone asphalt pavements; moreover, existing skid resistance studies predominantly emphasize single-factor analysis, resulting in an insufficient understanding of the evolution laws of skid and abrasion resistance under multi-factor coupling effects. On the other hand, the intrinsic relationship between pavement surface morphology changes and the degradation of skid and abrasion resistance has yet to be fully revealed.

On this basis, a laboratory accelerated abrasion test device integrating abrasion and real-time testing functions was self-developed in this study. Based on this device, laboratory-molded rut slab specimens were used to simulate asphalt pavement, and the evolution law of the long-term skid resistance and abrasion resistance of limestone asphalt pavement under the coupling effect of different temperatures and loading speeds was systematically investigated. Meanwhile, a high-resolution camera was used to capture the surface morphology of specimens at different abrasion stages, and the morphological change characteristics and their correlation with performance attenuation were analyzed. The research results can provide a theoretical basis for the rational application of limestone in the surface layer of asphalt pavement, and have certain engineering significance for reducing project costs and promoting the efficient utilization of local resources.

2 Materials and test methods

2.1 Materials

2.1.1 Asphalt

A 70# petroleum road asphalt was selected as the base asphalt. Its technical properties were tested in accordance with Standard Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering (JTG E20-2011), and the test results are presented in Table 1. All measured indices meet the requirements of the specification.

TABLE 1. Techincal indexes of 70# asphalt.

2.1.2 Aggregates

The technical indices of limestone coarse and fine aggregates were tested in accordance with *Test Methods of Aggregate for Highway Engineering* (JTG E42-2005). The test results are presented in Table 2 and 3, and all indices meet the requirements of relevant specifications.

TABLE 2. Limestone coarse aggregate technical specification.

TABLE 3. Testing results of basic technical indicators of limestone fine aggregate.

2.1.3 Mineral filler

Limestone mineral filler was adopted in this study. Its technical properties are presented in Table 4, all of which meet the requirements of the specification *Test Methods of Aggregate for Highway Engineering* (JTG E42-2005).

TABLE 4. Technical properties of limestone mineral filler.

2.2 Laboratory accelerated abrasion equipment

2.2.1 Equipment design and structure

The working principle of the laboratory abrasion testing machine and the accelerated abrasion device is shown in Figure 1. The equipment is mainly composed of a loading assembly, a control panel, a temperature and humidity regulation system, an electrical control box, a drainage outlet, and an air pump. The accelerated abrasion device includes a transmission assembly, a loading assembly, and a rotary loading platform. The transmission assembly and the loading assembly are connected by a reinforced connecting plate, and the loading assembly is fixed to the rotary loading platform through support columns.

The transmission assembly consists of a servo motor, a transmission belt, and a transmission shaft. The servo motor drives the transmission shaft via a synchronous belt. The loading assembly, from top to bottom, includes a specimen penetration displacement sensor, a wheel angle adjustment device, a pressure sensor, a lateral force sensor, a wheel bracket, a lifting cylinder, and a wheel. The rotary loading platform is composed of a steel plate track, a specimen mounting bracket, and a bottom plate. The steel plate track is fixed to the ground by bolts and serves as the test platform, which can accommodate four standard rut slab specimens with a size of 300 mm × 300 mm × 50 mm.

The tire loading motion is rotational, and the wheel angle can be adjusted outward within a range of 0°–20°. The control panel allows precise adjustment of the instrument parameters, including tire lifting, abrasion cycles, traveling speed, and applied load. Specifically, the load ranges from 0 to 1.2 MPa, the speed ranges from 0 to 60 km/h, the temperature ranges from −30 °C to 80 °C, and the humidity ranges from 0% to 95% RH, which can accurately reproduce the deterioration process of the abrasion resistance of asphalt pavement under the coupling effect of multiple factors. The device can accommodate four identical specimens simultaneously and adopts two types of tires with different materials, which are respectively used for pavement abrasion (abrasion resistance) and friction coefficient testing (skid resistance), so as to more accurately characterize the evolution of skid resistance and abrasion resistance under external loading. Meanwhile, the device can automatically collect friction coefficient and abrasion depth (used for calculating thickness loss rate) in real time, with high efficiency and accurate results. In addition, the device can not only conduct efficient abrasion and precise testing on prefabricated specimens in the laboratory environment, but also realize on-site real-time sampling, closely linking the research with actual road conditions and improving the reliability of the test results.

2.2.2 Equipment test steps

The test procedure of the device is as follows:

• Check whether all fasteners are secure, baffles are properly installed, and the tire tread is flat and free of damage. Verify that the temperature and humidity sensors and the real-time data display system function normally. Inspect the sealing of the protective window and confirm that the emergency stop button is sensitive and reliable to ensure test safety.
• Start the temperature and humidity control system and begin the test after the environmental conditions stabilize. Firmly mount the specimen on the fixture and remove surface dust to prevent specimen displacement or impurities from affecting the test.
• Manually rotate the tire for one full cycle and reset the abrasion depth and friction coefficient to zero. Set the applied load, operating speed, and number of abrasion cycles. Close the protective window and gradually start acceleration. Record the average abrasion depth and friction coefficient collected in real time by the device at the preset abrasion stages.
• Stop the device after completing the abrasion test. Take out the specimen after the temperature and humidity return to normal, and clean the device and residual debris.

2.3 Asphalt mixture mix design

According to the requirements of Standard *Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering* (JTG E20-2011), the Marshall test method was used in this study to conduct the mix design of asphalt mixtures and determine the optimal asphalt-aggregate ratio. The gradation curves of AC-13, AC-16 and AC-20 asphalt mixtures are shown in Figure 2. The optimal asphalt-aggregate ratios of AC-13, AC-16 and AC-20 asphalt mixtures are 4.65%, 4.42% and 4.21%, respectively.

2.4 Test methods

2.4.1 Thickness loss rate

In this study, the contact pressure between the abrasion tire and the pavement was set to 0.7 MPa. After setting the conditions including temperature, humidity, speed, and abrasion cycles, the abrasion test was carried out using the abrasion wheel. The formula for calculating the thickness loss rate of the specimen before and after abrasion is shown in Equation 1. The working diagram of the abrasion wheel is presented in Figure 3.

h = H 1 − H 2 L (1)

Where h is the thickness loss rate (mm/m), H₁ is the initial thickness of the specimen (mm), H₂ is the post-abrasion thickness of the specimen (mm), and L is the length of the abrasion section (m), which is fixed at 0.4021 m.

The thickness loss rate was measured using a built-in penetration displacement sensor installed on the test apparatus. After the specimen was secured and the measurement program was initiated, the sensor automatically acquired multi-point initial thickness readings, and the average value was designated as H₁. During the abrasion process, at each preset abrasion cycle interval, the sensor automatically captured the downward displacement of the wheel induced by progressive surface wear; this displacement was averaged over multiple measurement points to obtain H₂, from which the apparatus automatically calculated and output the thickness loss rate corresponding to that specific cycle number. The wear section length of 0.4021 m was determined through preparatory calibration experiments: the tire was driven through one complete revolution around four standard rutting slab specimens, the circular trajectory of the tire on the specimen surfaces was physically measured, and the average arc length was calculated using the circumference formula. This value was subsequently adopted as a fixed geometric parameter in the wear tests.

2.4.2 Friction coefficient

After a specified number of abrasion test cycles were completed, we replaced the wheel with a specialized friction test wheel (with a camber angle of 15° and a preset travel speed of 5 km/h) to measure and record the coefficient of friction (COF) under different abrasion cycle counts. The coefficient of friction is defined as the ratio of the tangential force (lateral friction force) to the normal force (vertical load), both of which are acquired in real time by the built-in sensors of the test apparatus. The centrifugal force generated by wheel rotation increases the lateral force measured by the apparatus, which in turn leads to an overestimation of the COF; thus, the friction component induced by centrifugal force must be eliminated. For this purpose, no-load tests were conducted under identical test conditions, and the corresponding lateral force was recorded. The COF component induced by centrifugal force, determined via this calibration method, was 0.109 under the conditions of 0.7 MPa and 5 km/h. The schematic diagram of the working principle of the abrasion wheel and the friction test wheel for different abrasion cycles is shown in Figure 4.

2.4.3 Acquisition and calculation of texture parameters

The surface morphology of the asphalt pavement was acquired using a 3D laser scanner, with the apparatus shown in Figure 5. This scanner, model 5010 C, is manufactured by Zoller + Fröhlich (Z + F, Germany), with a scanning field of view of 360° horizontally and 320° vertically. It is capable of real-time data transmission, blind spot analysis, and on-the-fly point cloud data registration during scanning. During the test, the specimen was placed vertically on a rigid test bench, and 3D laser scanning was performed on the surface of the specimen. Upon completion of scanning, the point cloud data and images acquired by the 3D laser scanner were imported into Trimble RealWorks 12.0 software (Trimble Inc., USA), and the target data were extracted for the calculation of pavement texture parameters.

R_a and R_q are core parameters for quantifying surface roughness, and they perform distinct functions in pavement texture analysis. R_a focuses on characterizing and evaluating the overall undulation of the surface, and is suitable for describing texture features across a large spatial scale. R_q has a higher sensitivity to minor local variations, and can accurately reflect the fine microstructure of the surface. The combined use of these two metrics enables a systematic analysis of pavement texture at multiple spatial scales, with a complete characterization of its complex properties (Chen et al., 2021; Qian et al., 2024). Therefore, R_a and R_q were selected as the indicators for pavement texture characterization in this study. The calculation methods for the texture parameters R_a and R_q are summarized in Table 5.

TABLE 5. Structure depth index test method.

2.4.4 Acquisition of specimen surface morphology

In this study, it was found that the surface morphological characteristics of the specimens were essentially consistent when the number of abrasion cycles reached 80,000 and 100,000. Furthermore, there was a significant difference in the degradation process of asphalt pavement surface morphology under low-temperature conditions compared with that under high-temperature conditions. Therefore, at an abrasion speed of 10 km/h, a high-resolution digital camera was used to capture the surface morphology of specimens subjected to 20,000, 40,000, 60,000 and 80,000 abrasion cycles under 0 °C and 60 °C conditions, respectively. This test was performed to analyze the evolution law of long-term skid resistance and abrasion resistance of limestone asphalt pavement, and to reveal the interaction mechanism between the macroscopic pavement performance and the evolution of pavement surface morphology.

3 Investigation on long-term skid resistance and abrasion resistance of limestone asphalt pavement

3.1 Thickness loss rate

The variation laws of the thickness loss rate of AC-13, AC-16 and AC-20 limestone asphalt specimens under different temperature and travel speed conditions are shown in Figure 6.

As shown in Figure 6, the thickness loss rate of AC-13, AC-16, and AC-20 limestone asphalt specimens all increases with the number of abrasion cycles, exhibiting a three-stage variation trend: slow growth in the initial phase, subsequent accelerated growth, and eventual stabilization. The underlying mechanism for this trend is as follows. Among specimens with different aggregate gradations, the AC-13 limestone asphalt specimen shows the highest thickness loss rate, which is mainly attributed to its higher fine aggregate content and relatively loose internal structure, resulting in weaker resistance to external loads (Guo et al., 2023). In contrast, AC-16 and AC-20 limestone asphalt specimens have a higher proportion of coarse aggregate, larger aggregate particle size, and more uniform asphalt film distribution, which enables the formation of a more stable skeleton structure and thus leads to a lower thickness loss rate (Zhou et al., 2023). This result indicates that asphalt mixture pavement with larger aggregate particle size and well-graded gradation has better abrasion resistance than mixture with small-size aggregate.

As shown in Figure 6, the thickness loss rate increases with increasing abrasion speed. Under the condition of 30 °C and 5 km/h, the maximum thickness loss rate of the specimen is 15.42 mm/m; at 30 °C and 10 km/h and 15 km/h, the maximum thickness loss rates are 16.91 and 19.4 mm/m, respectively. This is because the kinetic energy imparted by the tire to the pavement surface increases significantly with increasing abrasion speed, leading to greater vertical impact forces and tangential friction forces generated during both the rolling and sliding motions of the tire. These forces act not only on the specimen surface but also are transmitted into the internal aggregate structure of the specimen, subjecting the material to higher instantaneous stress and cumulative damage. Accordingly, the micro-spalling between aggregates and the plastic deformation of the asphalt binder are further aggravated, which significantly accelerates the growth rate of the thickness loss rate. Furthermore, under high abrasion speed, the stress borne by the pavement material may exceed its bearing capacity, while the viscoelasticity of the asphalt binder is insufficient to effectively buffer the external loads. This makes the interlocking structure between aggregates more vulnerable to damage and aggravates the abrasion of the pavement (Majid and Ismael, 2019).

At an abrasion speed of 10 km/h, the maximum thickness loss rates of the limestone specimens after 10 × 10⁴ abrasion cycles at 0 °C, 30 °C, and 60 °C are 18.65, 16.91, and 21.89 mm/m, respectively. This is attributed to the fact that under high-temperature conditions, the viscosity of the asphalt binder decreases and its bonding strength weakens, making aggregates more susceptible to stripping. Consequently, the thickness loss rate reaches its maximum value under the high-temperature condition (Akraym et al., 2023; Zhang et al., 2025). Under low-temperature conditions, although the bonding strength between asphalt and aggregates is improved, the asphalt material becomes more brittle and the pavement is more susceptible to cracking. At room temperature, the asphalt has a moderate viscosity, which can not only ensure sufficient adhesion with aggregates but also allow a certain degree of deformation, thus leading to the lowest thickness loss rate. This demonstrates that the pavement structure has the most stable and optimal abrasion resistance at room temperature (Bai et al., 2024).

As shown in the figure, the difference in thickness loss rate between the AC-16 and AC-20 limestone specimens is 1.99 mm/m at 0 °C and 10 km/h, but decreases to 1.24 mm/m at 60 °C and 10 km/h, indicating that the gap narrows considerably under high-temperature conditions. This may be because high temperatures degrade the performance of the asphalt binder, loosen the aggregate skeleton structure, and consequently reduce the internal structural stability of the mixture. In addition, under the condition of 30 °C–5 km/h, the difference in thickness loss rate between AC-13 and AC-16 specimens is smaller than that between AC-16 and AC-20 specimens under the condition of 30 °C–10 km/h, and is far smaller than the difference under the condition of 60 °C–10 km/h. These results further confirm that the abrasion resistance of pavement is affected not only by vehicle-related factors, environmental conditions, and aggregate particle size, but also more significantly by the coupling effect of multiple factors.

3.2 Coefficient of friction

The variation laws of the coefficient of friction for AC-13, AC-16, and AC-20 limestone asphalt specimens under different temperature and abrasion speed conditions are shown in Figure 7.

As shown in Figure 7, during the initial abrasion stage, the friction coefficients of the AC-13, AC-16, and AC-20 limestone specimens all decreased noticeably, primarily due to the compaction of the asphalt film under tire loading, the gradual blunting of aggregate edges, and the progressive densification of aggregate particles under load (Chen et al., 2025). At 30 °C and 5 km/h, both the AC-13 and AC-20 limestone specimens exhibited an increase in friction coefficient during the 0–20,000 abrasion cycles. Specifically, the friction coefficient of the AC-13 specimen increased from 0.281 to 0.292 after 10,000 cycles, while that of the AC-20 specimen increased from 0.307 to 0.313 after 20,000 cycles. At 0 °C and 10 km/h, the AC-13 specimen exhibited an increase from 0.302 to 0.308 after 5,000 abrasion cycles. This is because at low abrasion speeds, the shear action of the tire on the asphalt film is relatively mild. As the asphalt film is gradually worn away, the micro-texture of the originally covered coarse aggregates becomes exposed, increasing the micro-roughness of the surface and thereby temporarily enhancing the friction coefficient. With continued abrasion, the aggregate edges are progressively polished, and the friction coefficient subsequently decreases. At low temperatures, the asphalt film is more susceptible to brittle micro-cracking and flake-like spalling under tire rolling. Since the asphalt film of the AC-13 mixture is thinner, it is more readily worn off, leading to gradual exposure of the coated limestone aggregate edges and thus a rapid increase in the friction coefficient within a short period. As the abrasion speed increases, this rising trend is weakened, but the AC-20 specimen still exhibits a certain degree of increase at relatively high speeds. Furthermore, at 30 °C and 15 km/h, surface damage occurred on the AC-13 specimen at 80,000 abrasion cycles. The tire generated additional mechanical interlocking with the damaged area, elevating the friction coefficient from 0.219 to 0.222. In contrast, the AC-16 and AC-20 specimens did not exhibit this phenomenon owing to their superior wear resistance.

With the increase of abrasion speed, the difference in COF among the three types of specimens gradually decreased, as the surface texture of different specimens tended to be consistent during the abrasion process. In the late abrasion stage, the decline rate of the COF slowed down significantly and tended to stabilize (Sun et al., 2022a). Furthermore, the COF of AC-20 limestone asphalt specimens was significantly higher than that of AC-13 specimens, which is mainly due to its high coarse aggregate content, forming more prominent microscopic asperities and interlocking structures, and thus generating stronger mechanical interlocking and friction effects. The aggregate gradation of AC-16 limestone asphalt specimens is between that of AC-13 and AC-20, so its COF is also at an intermediate level.

Temperature also has a significant effect on friction performance. As shown in Figure 7f, at an abrasion speed of 10 km/h, after 100,000 cycles, the friction coefficient loss rates of the limestone specimens at 0, 30, and 60 °C were 15.23%–22.10%, 23.21%–26.06%, and 24.69%–31.14%, respectively, indicating that the friction coefficient loss rate increases with increasing temperature. This is because at low temperatures, the asphalt binder exhibits a relatively high stiffness modulus, and the bonding strength between the aggregate and the asphalt is well maintained. The deterioration of the pavement surface micro-texture is dominated by mechanical polishing, which proceeds relatively slowly, resulting in a lower friction coefficient loss rate. When the temperature increases to 30 °C, the asphalt gradually enters a viscoelastic state, and the bonding strength of the binder decreases. Under repeated tire loading, the aggregates are more prone to micro-displacement and polishing, leading to a corresponding increase in the friction coefficient decay rate. Under high-temperature conditions, the asphalt binder softens significantly and its viscosity drops markedly, resulting in a considerable reduction in its ability to confine the aggregates. Consequently, aggregate particles are more susceptible to stripping, migration, and edge polishing. Meanwhile, high temperatures intensify the plastic deformation and compaction effects on the pavement surface, further accelerating the degradation of surface texture and resulting in the highest friction coefficient loss rate. In addition, at 0 °C, the friction coefficients of the three specimens are generally higher and show pronounced differences among them, which is attributed to the more stable surface microstructure under low-temperature conditions (Zhang et al., 2023). When the temperature rises to 30 °C, the asphalt enters a viscoelastic state, and its partial fluidity weakens the influence of aggregate morphology on friction performance, thereby reducing the differences in friction coefficient among the specimens. When the temperature increases further to 60 °C, the asphalt becomes markedly softened, and the contribution of aggregate particle size to friction performance is further weakened. As a result, the differences in friction coefficient among the three specimens are significantly smaller than those at 0 °C, but are still slightly larger than those observed at 30 °C.

3.3 Texture parameters

The variation laws of the texture parameters of AC-13, AC-16, and AC-20 limestone asphalt specimens under different temperature and abrasion speed conditions are shown in Figure 8.

As can be seen from Figure 8, the R_a and R_q values of AC-20 limestone asphalt specimens were significantly higher than those of AC-13 specimens, but showed a minor difference from those of AC-16 specimens. This phenomenon is mainly attributed to the high content of coarse aggregate in AC-20 specimens, which forms more prominent and abundant microscopic asperities on the specimen surface, leading to more significant surface undulation and thus higher R_a and R_q values. In contrast, the surface of AC-13 limestone asphalt specimens is relatively smooth with fewer microscopic asperities, so its R_a and R_q values are lower than those of AC-20 specimens (Sun et al., 2024). AC-16 limestone asphalt specimens contain a certain proportion of coarse aggregate to form surface undulation, while the fine aggregate ensures a certain degree of flatness, so there is little difference in surface roughness between AC-16 and AC-20 specimens. Furthermore, AC-20 limestone asphalt specimens exhibited stronger resistance to microstructure damage in the early abrasion stage, so the decline rate of its R_a and R_q values was relatively slow. By contrast, the limited microscopic asperities on the inherently smooth surface of AC-13 specimens were easily polished, resulting in a rapid decrease in R_a and R_q values. Meanwhile, due to the relatively weak resistance to external loads of AC-13 specimens, obvious surface distress was more likely to occur in the late abrasion stage, leading to a sudden increase in R_a and R_q values.

Under the same number of abrasion cycles and specimen type, the decrease in R_a and R_q values became more pronounced with the increase of abrasion speed, and the curves showed a steeper downward trend at higher speeds. With the increase of abrasion cycles, the differences in R_a and R_q among the three types of specimens gradually decreased. This is because the abrasion process weakened the differences in surface microstructure caused by aggregate particle size, and made the surface roughness of different specimens gradually converge. This variation law is consistent with the change trend of the coefficient of friction. At 0 °C, the R_a and R_q values of the three types of specimens were generally high, while these values decreased significantly at 60 °C. This is because the surface microstructure of the specimens was relatively stable under low-temperature conditions, while the specimen surface was more prone to deformation and abrasion during the high-temperature abrasion process.

As can be seen from the figure, temperature variations also exert a certain influence on the evolution of texture parameters. Under the condition of 0 °C and 10 km/h, after abrasion, the maximum reductions in R_a and R_q of the limestone specimens were 29.83% and 26.56%, respectively; whereas under 60 °C and 10 km/h, the maximum reductions in R_a and R_q were 40.00% and 32.30%, respectively. This indicates that high-temperature conditions significantly accelerate the degradation of pavement surface texture. The reason is that at low temperatures, the asphalt binder is in a high-elastic state with strong bonding with the aggregates, and the surface microstructure is relatively stable; thus, texture degradation mainly manifests as slow mechanical polishing. In contrast, at high temperatures, the asphalt binder softens markedly, its viscosity decreases, and its ability to confine the aggregates is weakened. Consequently, aggregate particles are more susceptible to displacement, stripping, and polishing under repeated tire loading. In addition, high temperatures intensify the plastic deformation and compaction effects on the pavement surface, thereby accelerating both the rate and the extent of surface texture degradation.

3.4 Surface morphology

3.4.1 Abrasion surface morphology

The evolution laws of the abrasion surface morphology of AC-13, AC-16, and AC-20 limestone asphalt specimens under the condition of 60 °C and 10 km/h are shown in Figure 9.

As shown in Figure 9, AC-13, AC-16, and AC-20 limestone asphalt specimens exhibit significantly different abrasion characteristics under different numbers of abrasion cycles. For AC-13 specimens, due to the low aggregate hardness and high fine aggregate content, aggregate breakage occurred after 20,000 abrasion cycles, generating a large number of fine abrasive particles and causing damage to the internal structure, which deteriorated the abrasion resistance of the specimens (Yu et al., 2023b). When the number of abrasion cycles increased to 40,000–60,000, the continuous accumulation of broken particles and repeated local loading from the tire led to the formation of irregular pits on the surface, further weakening the structural stability. After 80,000 abrasion cycles, the sustained load caused obvious surface plastic deformation and visible rutting, resulting in a significant decrease in abrasion resistance. In contrast, for AC-16 specimens, aggregate breakage and damage to the internal skeleton structure occurred after 20,000 abrasion cycles, which weakened the interlocking effect between aggregates and reduced the abrasion resistance. When the abrasion cycles reached 40,000, the lateral tire pressure caused local edge collapse, further reducing the edge durability. At 60,000 cycles, continuous aggregate breakage and edge collapse led to the formation of columnar pits and transverse cracks on the surface, accompanied by slight plastic deformation. After 80,000 abrasion cycles, the rutting was further deepened and expanded, indicating further degradation of its abrasion resistance.

For AC-20 specimens, the large particle size of coarse aggregates leads to the formation of large voids between coarse aggregates. Under the thermo-mechanical coupling effect, the ability of the asphalt binder to fill these voids is weakened, making the coarse aggregates more prone to spalling and increasing particle loss (Yu et al., 2023b). When the abrasion cycles reached 40,000, the lateral tire pressure caused edge collapse; at 60,000 cycles, the voids between aggregates were further expanded and formed columnar pits, while stress concentration promoted crack development and aggravated structural damage. After 80,000 abrasion cycles, long-term high temperature and cyclic loading caused local subsidence and reduced the structural bearing capacity, further decreasing the abrasion resistance. The variation of the specimen thickness loss rate further confirms that AC-13 specimens show the most significant overall deterioration, with a high growth rate and amplitude of thickness loss rate; while AC-16 and AC-20 specimens are mainly characterized by local damage. This indicates that aggregate particle size and skeleton structure have a critical impact on the abrasion resistance of asphalt pavement. Asphalt mixture with high coarse aggregate content and large particle size can transfer loads more effectively through aggregate interlocking, and better resist tire abrasion under thermo-mechanical coupling, resulting in a slower growth of thickness loss rate. On the contrary, although the mixture with dense fine aggregate gradation has a uniform initial structure, it is more prone to particle breakage and abrasive particle generation, with weaker shear and impact resistance, thus accelerating the overall deterioration of abrasion resistance (Li et al., 2024).

The evolution laws of the abrasion surface morphology of AC-13, AC-16, and AC-20 limestone asphalt specimens under the condition of 0 °C and 10 km/h are shown in Figure 10.

As shown in Figure 10, the surface morphology of AC-13, AC-16, and AC-20 limestone asphalt specimens under the condition of 0 °C and 10 km/h gradually degrades with the increase of abrasion cycles. At 20,000 abrasion cycles, all specimens mainly exhibited aggregate breakage accompanied by particle spalling. During 40,000–60,000 abrasion cycles, local stress concentration was prone to occur around the spalled particles, while the debris generated by aggregate breakage aggravated the friction between the tire and the specimen surface, accelerating the abrasion process. After 80,000 abrasion cycles, transverse cracks and surface spalling appeared on the specimen surface, which further aggravated the structural damage and led to a significant decrease in abrasion resistance.

By comparing the evolution of surface morphology under 0 °C and 60 °C conditions, significant differences in failure mechanisms can be identified. Under the 0 °C condition, the failure was relatively uniform, mainly manifested as aggregate breakage and particle spalling, with slow and moderate accumulation of structural damage. In contrast, under the 60 °C condition, the failure mode was more complex, including aggregate breakage, edge collapse, particle spalling, rutting, local subsidence, and crack formation. The accelerated degradation under high temperature is attributed to the reduced viscosity and weakened bonding strength of the asphalt binder, as well as the higher susceptibility to deformation of the pavement structure, which further aggravates the structural damage.

3.4.2 Skid resistance surface morphology

The evolution laws of the skid resistance surface morphology of AC-13, AC-16, and AC-20 limestone asphalt specimens under the condition of 60 °C and 10 km/h are shown in Figure 11.

As shown in Figure 11, for AC-13 limestone asphalt specimens, aggregate migration, coarse aggregate exposure, and asphalt film spalling occurred after 20,000 abrasion cycles, indicating that the bonding between asphalt and aggregates began to be affected in the early stage, and part of the coarse aggregates had been exposed. At 40,000 abrasion cycles, in addition to the abrasion of coarse aggregates, part of the fine aggregates also began to be exposed, and the degree of abrasion was further aggravated. The main feature at 60,000 abrasion cycles was the continuous abrasion of coarse and fine aggregates; as the abrasion proceeded, the aggregate surfaces were further smoothed. When the abrasion reached 80,000 cycles, the coarse and fine aggregates were severely polished, and the specimen surface tended to be smooth.

For AC-20 and AC-16 limestone asphalt specimens, the surface morphology was basically consistent. At 20,000 abrasion cycles, asphalt film spalling and coarse aggregate exposure occurred; at 40,000 cycles, the surface coarse aggregates were abraded, accompanied by a small amount of fine aggregate exposure, which led to a rapid decrease in the coefficient of friction and texture parameters. At 60,000 cycles, the coarse aggregates were polished, and a large area of fine aggregates was exposed. At this stage, the friction was mainly borne by the fine aggregates, and the decline rate of COF and texture parameters slowed down. At 80,000 cycles, the coarse and fine aggregates were severely abraded, the surface suffered serious abrasion, the edges of the aggregates were almost completely blunted, and the surface became smooth, with the COF and texture parameters basically tending to stabilize.

By comparing the evolution trends of the surface abrasion morphology of the three types of specimens, it can be found that AC-13 limestone asphalt specimens exhibited more significant abrasion deterioration and a faster abrasion rate at each stage. This is mainly because the aggregate particle size of AC-13 specimens is relatively small, and the problem of early fine aggregate exposure is more prominent under the same external load, which accelerates the abrasion rate of the specimens (Sun et al., 2022b). In the late abrasion stage, the overall abrasion is more severe due to the weak internal skeleton structure.

The evolution laws of the skid resistance surface morphology of AC-13, AC-16, and AC-20 limestone asphalt specimens under the condition of 0 °C and 10 km/h are shown in Figure 12.

As can be seen from Figure 12, under the condition of 0 °C and 10 km/h, all three types of specimens underwent an abrasion process consisting of asphalt film spalling, exposure of coarse and fine aggregates, and smoothing of aggregate surfaces. Compared with the evolution trend of abrasion surface morphology under 60 °C, the specimens at high temperature (60 °C) exhibited a faster morphological degradation rate, more severe damage, and more complex apparent characteristics. The fundamental reason for this phenomenon is that high temperature weakens the bonding strength of the asphalt binder, leading to a reduction in the overall structural stability of the specimens (Tang et al., 2023). Under the continuous unidirectional rolling of the tire, the specimens are more prone to plastic deformation (Kezhen et al., 2022), resulting in a larger decline rate and amplitude of the coefficient of friction (COF) and texture parameters during the abrasion process.

In contrast, under the low-temperature condition of 0 °C, the specimen surface was dry, and no greasy viscous film formed by asphalt flow was observed. The abrasion process was characterized by the gradual spalling of the asphalt coating, exposure of coarse and fine aggregates, and staged edge blunting of the aggregates. This staged mechanism of “aggregate exposure followed by polishing” led to a non-monotonic variation of the texture parameters: they were relatively stable in the early stage, then decreased slowly or even increased slightly, dropped significantly in the late stage, and exhibited a markedly enhanced overall fluctuation.

4 Conclusion

Based on a self-developed laboratory abrasion test apparatus, this study systematically analyzed the coupling effect of aggregate gradation (AC-13, AC-16, AC-20 limestone asphalt specimens), abrasion speed and temperature on the long-term evolution of asphalt pavement skid resistance and abrasion resistance. The degradation mechanism of pavement surface morphology under different temperatures was discussed, and the intrinsic correlation between surface morphology evolution and performance deterioration was revealed. The main conclusions are drawn as follows.

• The custom-built accelerated wear testing system can effectively reproduce the coupled effects of multiple factors, including temperature, loading speed, and aggregate particle size, thereby enabling reliable evaluation of the long-term skid and wear resistance of asphalt pavements.
• The long-term performance evolution of limestone asphalt pavement is synergistically governed by aggregate particle size, temperature, and loading speed. Skid resistance exhibits a dual-mode behavior, characterized by either a “rapid decay–stabilization” pattern or a “brief increase–continuous decline” pattern. In contrast, wear resistance consistently follows a three-stage pattern: “slow decay–rapid decay–stabilization.” Considering the synergistic effects of environmental conditions, traffic loading, and material properties can facilitate a more accurate representation of the long-term wear evolution of the pavement, thereby providing a reliable scientific basis for pavement design optimization and material selection. (2) The long-term performance evolution of limestone asphalt pavement is synergistically governed by aggregate particle size, temperature, and loading speed. Skid resistance exhibits a dual-mode behavior, characterized by either a “rapid decay–stabilization” pattern or a “brief increase–continuous decline” pattern. In contrast, wear resistance consistently follows a three-stage pattern: “slow decay–rapid decay–stabilization.” Considering the synergistic effects of environmental conditions, traffic loading, and material properties can facilitate a more accurate representation of the long-term wear evolution of the pavement, thereby providing a reliable scientific basis for pavement design optimization and material selection.
• Aggregate particle size governs the damage mode through its effect on skeleton stability: the AC-13 mixture is more susceptible to overall structural instability and accelerated deterioration, whereas the AC-20 mixture is dominated by localized damage and thus maintains stronger durability. However, under high-temperature conditions, asphalt binder bond failure significantly weakens the particle size effect, narrowing the performance differences among different gradations.
• Temperature is a critical environmental factor determining the failure mode and degradation rate of the pavement. At low temperatures, the damage is characterized by a single dominant mechanism, i.e., brittle aggregate fracture and slow polishing, resulting in a gradual performance decline. At high temperatures, in contrast, plastic flow, multiple damage types (e.g., rutting, collapse, and cracking), and additional compaction effects are induced, leading to sharp deterioration in performance. This temperature-controlled mechanism explains the intrinsic link between surface morphology evolution and macroscopic performance degradation.

This study employed a custom-built accelerated wear apparatus to investigate the evolution of long-term skid resistance and wear resistance of limestone asphalt mixtures under varying temperatures, loading speeds, and gradation conditions. It should be noted that this study was confined to the skid and wear resistance of limestone asphalt pavements; direct comparisons with hard rocks such as basalt were not conducted, and the mechanisms through which limestone aggregates intrinsically influence the skid resistance and wear behavior of asphalt pavements were not clearly elucidated. Future research will build upon this testing system to systematically conduct comparative tests between limestone and other hard rocks under identical test conditions, quantify the differences in skid resistance and wear resistance across different lithologies, and explore the use of composite aggregates or modified asphalt to enhance the service performance of limestone pavements.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Chen Li, Inner Mongolia University, China

Reviewed by: Chao Zhang, Changsha University of Science and Technology, China

Xiuze Fan, Washington State University, United States