Study on Skid Resistance of Asphalt Pavements Under Macroscopic and Microscopic Texture Features: A Review of the State of the Art

Abstract

Pavement skid resistance is one of the most important factors affecting road safety, and pavement texture morphology significantly influences this property. Therefore, it is crucial to investigate the relationship between pavement texture and skid resistance. This article provides an overview of recent research advancements in key areas, including the anti-skid mechanisms of asphalt pavements, factors affecting pavement anti-skid performance, methods for characterising and evaluating pavement anti-skid performance, and the relationship between the macroscopic and microscopic texture of pavements and their anti-skid performance. Based on a comparative analysis of the intrinsic mechanical interactions between asphalt pavements and rubber tyres, it was determined that the surface texture characteristics of the asphalt pavement are the most critical factor influencing its anti-skid performance. These include both macroscopic and microscopic texture parameters, which, together with the service environment, collectively influence the pavement’s anti-skid performance. The existing texture characteristics, based on the anti-skid performance of asphalt pavements, as detected by various methods and evaluated using established models, are summarised here. Finally, this article discusses the relationship between texture characteristic parameters and asphalt pavement anti-skid performance from both macro- and microtexture perspectives. This synthesis serves as a valuable reference and basis for further research and development in enhancing asphalt pavement skid resistance.

1. Introduction

Road transport serves as the principal modality for passenger and freight mobility in the European Union, supported by mature transportation infrastructure. Official statistics revealed that in 2017, the EU’s road network spanned 6.1 million kilometres, with 73,000 km constituting dedicated motorway infrastructure. Paradoxically, this sectoral expansion has coincided with escalating safety concerns [1]. Compelling data from the European Transport Safety Council (ETSC, 2023 Annual Report) document 20,418 road traffic fatalities in 2023. Of these, approximately 4371 had accidents with causes related to the deterioration of the pavement’s skid resistance. Among critical determinants, pavement skid resistance emerges as a pivotal engineering parameter influencing collision risk mitigation. Approximately 25% of all global aviation accidents are associated with runway excursions, 35% of which result directly from inadequate surface friction. Wet or waterlogged runways are susceptible to hydroplaning, whereas snowy or icy conditions are characterised by extremely low friction coefficients, often leading to a total loss of braking capability. For instance, the United States Air Force experienced as many as 20 such incidents in 1973 alone, highlighting the severe safety risks posed by insufficient skid resistance during aircraft landing and take-off operations. This causal nexus necessitates mechanistic understanding of tyre–pavement friction dynamics, as enhanced skid resistance directly correlates with improved crashworthiness—a fundamental prerequisite for advancing transportation system safety.

The skid resistance of a road can be defined as the road surface’s ability to prevent tyres from sliding in the direction of travel, which is quantified by the coefficient of friction [2]. A review of the literature indicates that, in addition to subjective factors, skid resistance depends on several parameters: driver behaviour, vehicle performance, road surface roughness, and environmental factors (such as temperature, rain, snow, and pollutants) [3]. Consequently, enhancing skid resistance in pavements has become a critical research focus in the field of pavement safety [4].

The Permanent International Association of Road Congresses (PIARC) has established a systematic classification of pavement texture, dividing it into four distinct categories: microtexture, macrotexture, megatexture, and unevenness. This classification is based on the wavelength and amplitude characteristics of surface irregularities. Over the years, a substantial body of research has been devoted to the investigation of pavement texture, with numerous studies examining the topic from diverse scientific and engineering viewpoints. In the evaluation of pavement texture, researchers have investigated the statistical geometric features, spectral features, fractal features, and multifractal features of pavement texture. Regarding the measurement of pavement texture, scholars in this field have explored methodologies for both the direct and indirect assessment of pavement macrotexture. Asphalt pavement texture exhibits a clear random distribution and complex variations. Consequently, characterising pavement skid resistance based on the macroscopic and microscopic texture characteristics of the pavement remains a challenging and contentious area of research both domestically and internationally.

The core innovation of this study lies in two aspects: (1) a systematic review of the synergistic influence mechanism of both macroscopic and microscopic textures of asphalt pavements on skid resistance, and (2) an investigation into the characterisation and evaluation systems for pavement texture and skid resistance. These contributions provide theoretical foundations and practical guidance for the design and maintenance of skid-resistant pavements.

2. Research on Anti-Skid Mechanisms of Asphalt Pavement

2.1. Mechanical Mechanism Between Tyres and Road Surfaces

2.1.1. Van Der Waals Forces Between Tyres and Road Surfaces

The skid resistance of the road surface is defined as the resistance generated by the sliding of tyres on the road surface when a moving vehicle is subjected to braking [5]. When a vehicle is driven normally, the rubber molecules of the tyre and the road surface molecules are in very close proximity resulting in the weak attraction generated by transient charge polarisation, which directly affects the anti-skid performance [6]. In 1963, Kummer et al. [7] conducted a pertinent study on the interaction force between the tyre and the road surface. They proposed that the adhesion force between the tyre and the road surface can be attributed primarily to the adhesion force and hysteresis force. The road surface adhesion force can be expressed by Equations (1) and (2). From the perspective of the tyre–road contact area, the contact area can be divided into an adhesion area and a sliding area. But Zheng et al. [8] considered that friction (adhesion) has three components: attachment, hysteresis, and ploughing. As shown in Figure 1, with increasing slip rate, the shear stress within the tyre contact patch continues to increase, and when the elastic deformation limit of the rubber molecular chain is exceeded, the adhesion region is forced to shift to the sliding region (macroslip), resulting in the expansion of the sliding region and the contraction of the adhesion region. The maximum friction factor may occur at any point in the contact region while the vehicle is moving, particularly when the slip rate is between 7% and 25%. Therefore, for the analysis of tyre–road contact, the tyre–road adhesion characteristics rather than friction characteristics should be considered. Zhao and Wang et al. [9,10,11,12] investigated the Van der Waals forces between tyres and rough road surfaces. Grosch concluded that the adhesion between the road surface and the tyre results from shear action on the contact surface, especially on smooth, clean surfaces. The hysteresis force is correlated with the attenuation of rubber damping energy. The vehicle brakes most effectively when the longitudinal adhesion force reaches its peak. When the tyre is in partial slip, the tyre experiences an elastic slip phenomenon, where the slip area expands to cover the entire contact area. The friction applied to the tyre is known as the limiting adhesion force, and its magnitude is equal to the friction when the entire contact area is sliding. At this point, Coulomb’s law no longer applies. The tyre–pavement contact region is divided into two distinct domains: an adhesion domain and a sliding domain. In road engineering, a friction factor is typically used to quantify the Van der Waals forces between the road surface and the tyre [13]. Friction factors can be classified as either rolling or sliding, depending on the state of vehicle motion and braking. The friction factor is influenced by various factors, including the characteristics of the road surface, the performance of the tyre, the operating parameters of the vehicle, and environmental factors [13,14].

$$F_a={\displaystyle \sum }_{i=1}^n{\tau }_i\times A_i={\displaystyle \sum }_{i=1}^n\phi \times {\sigma }_i\times A_i$$

(1)

$$F_{\mathrm{h}}=c{\displaystyle \sum }_{i=1}^n\left(E_{c_i}-E_{e_i}\right)$$

(2)

where $F_a$ is the adhesion force, ${\mathrm{F}}_{\mathrm{h}}$ the drag force of the tyre, ${\tau }_{\mathrm{i}}$ is the interfacial shear strength, $\mathrm{n}$ is the effective number of micro-bumps, $\phi$ is the adhesion coefficient between the tyre and the road surface, ${\sigma }_{\mathrm{i}}$ is the normal contact stress on the surface of the first i microbump, $\mathrm{c}$ is the constant of proportionality, and ${\mathrm{E}}_{{\mathrm{c}}_{\mathrm{i}}}{-\mathrm{E}}_{{\mathrm{e}}_{\mathrm{i}}}$ is the tyre energy loss on the surface of a single micro-bump.

2.1.2. Adhesion Between Tyre and Road Surface

In the context of tyre and pavement material research, rubber, as a viscoelastic material with a low modulus of elasticity, demonstrates an aptitude for recuperating deformation due to the intrinsic friction of the rubber material. This enables the tyre to adhere to the road surface, with the contact area exhibiting a nonlinear relationship. The adhesion characteristics of asphalt pavement, which is also a viscoelastic material sensitive to temperature, become evident when the tyre is in close contact with the road surface, resulting in intermolecular adhesion between the two. Asphalt pavement is also a temperature-sensitive viscoelastic material. Consequently, intermolecular adhesion occurs between the tyre and the pavement when the two are in close contact [15]. The friction test between the tyre and the road surface revealed the presence of minute particles of road material on the surface of the tyre, as well as the observation of similar particles of tyre rubber material on the surface of the road. This evidence indicates that adhesion does exist between the tyre and the road surface. During the test, when the tyre slips and slides over the aggregate surface, the tyre undergoes an adhesion deformation process. This process consists of a compression phase on the left side of the rubber particles and a recovery phase on the right side of the rubber particles. It is not possible to transfer equal amounts of energy during the compression and recovery phases. Consequently, the dissipated energy is released in the form of heat generation [16], as illustrated in Figure 2. For this part of energy dissipation, Srirangam found the equation for the energy dissipation of the cell due to the adhesion between tyre and road surface as well as due to heat generation [17], while Wagner et al. [18] developed a model for the physical quantities of the adhesion and hysteresis forces between tyres and the road surface.

2.1.3. Microcutting of Small-Sized Microconvex Bodies on Pavements

When the contact stress between tyre and road surface is investigated, it can be observed that when a tyre of a highly loaded vehicle is in contact with the sharp raised particles on the road surface, the raised particles tend to lead to stress concentration on the surface of the tyre at a localised area [19]. When the concentrated stress is greater than the fracture strength of the rubber material, the microconvex body will have a microcutting effect on the tyre surface, a phenomenon similar to the ploughing action in metal tribology. Microcutting of the tyre surface by the microconvex body is subject to the resistance of the tyre surface [20], a process that is also part of the friction between the tyre and the road surface. The microcutting effect of the microconvex body on the tyre surface was also further confirmed by researchers using scanning electron microscopy to observe the surface of tyres that had undergone friction tests, as shown in Figure 3 [21]. It has been found that the friction due to the microcutting action of the microconvex bodies of the road surface is related to the material properties of the road surface and the tyre, as well as the density, size, and distribution pattern of the microconvex bodies. The microcutting friction can be calculated by Equation (3).

$${\mathrm{F}}_{\mathrm{c}}{=\mathrm{K}}_3{\mathrm{htg}(\delta )\mathrm{T}}_{\max }{=\mathrm{K}}_3{\mathrm{K}}_4{\left|{\displaystyle \frac{\mathrm{P}}{\mathrm{tg}\frac{\theta }{2}}}\right|}^{{\displaystyle \frac{1}{2}}}{\mathrm{tg}(\delta )\mathrm{T}}_{\max }$$

(3)

where ${\mathrm{K}}_3{,\mathrm{K}}_4$ are constants, N is the load normal to the peak of the microconvex body, $\theta$ is the simplified cone top angle of the microconvex body, h is the depth of the microconvex body piercing into the surface of the tyre, and ιgδ is the tangent modulus of the rubber.

The mechanical mechanism between tyre and pavement can be broadly categorised into three main aspects [22]:

• Van der Waals force action between tyre and pavement.
• Adhesion between tyre and pavement.
• Microcutting action of small-sized microconvex bodies on the pavement.

2.2. Factors Affecting the Skid Resistance of Asphalt Pavements

Aggregate characteristics, grading type, nature of asphalt, and asphalt dosage can reflect the macro- and microtexture characteristics of asphalt pavements. Among the main factors affecting the anti-skid performance of pavements are the macroscopic and microscopic texture characteristics of pavements.

2.2.1. Influence of Aggregate Properties on Skid Resistance of Asphalt Pavements

The skid resistance of asphalt pavements is primarily determined by the intrinsic properties of the aggregate particles. Both coarse and fine aggregates employed in asphalt mixtures significantly influence the surface friction performance. During the early stage of pavement service life, the asphalt binder partially covers the aggregate surfaces, which modulates the tyre aggregate interaction and subsequently affects the development of skid resistance [23]. Although the binder is present, aggregates continue to constitute the primary contact interface between the pavement and vehicle tyres and therefore play a pivotal role in the generation of friction forces. In order to achieve optimal skid resistance in pavement construction, it is essential to select coarse and fine aggregates of the highest possible quality. Masad et al. [24] found that mineral specification affects the macroscopic tectonic depth of the pavement and stated that high quality stones such as sandstone, basalt, andesite, and granite have better skid resistance, abrasion resistance, and impact resistance. The aggregate surface texture structure represents the primary component of the pavement microtexture, characterised by a horizontal wavelength of less than 0.5 mm and an amplitude of 1 to 500 μm. The formation of microtexture is therefore influenced by two main factors: the type of mineral composition of aggregate particles and the content of mineral components. The microtexture of aggregate is the basis of good anti-skid performance of asphalt pavement [25,26,27,28,29].

In terms of the mineral composition of aggregate particles, aggregates with hard crystal structures and strong bonding and interlocking between crystals generally provide good wear resistance [30,31,32]. The assessment of aggregate mineral composition requires the application of microstructural analytical techniques, including optical microscopy and scanning electron microscopy (SEM) [31]. According to previous studies, an optimal content of hard minerals—generally falling within the range of 50% to 70%—can significantly enhance skid resistance. Moreover, aggregates with an average crystal grain size of approximately 200 μm, combined with angular-shaped hard mineral grains, are typically regarded as the most favourable for improving frictional pavement performance. Aggregate abrasive resistance is the capacity of an aggregate to sustain the microtexture of the aggregate surface following exposure to a multitude of abrasive and shear forces exerted by traffic loads.

2.2.2. Influence of Gradation Type of Mix on Skid Resistance of Asphalt Pavements

Asphalt mixture is composed of coarse aggregate, fine aggregate, filler, and asphalt material in accordance with a fixed proportion of a visco-elastic composite material. A reasonable mixture grading allows for the aggregate particles to have a beneficial embedded role and internal friction resistance. This, in turn, ensures that the asphalt pavement can maintain its design strength under the action of vehicular loads. The grading type of the mixture has a significant impact on the structure between coarse and fine aggregate particles, which is also known as macrotexture (horizontal wavelength of 0.5–50 mm, amplitude of 0.1–20 mm). The formation of macrotexture is primarily influenced by the following factors [33]: the size, shape, and distribution of coarse and fine aggregates; the nominal maximum particle size of the aggregates; and the use of pavement construction in the construction methods are the main factors that affect the grading type of the mix.

Different gradation types of mixes can have different effects on the skid resistance of pavements. For example, Chen et al. [34] demonstrated that the proportion of coarse aggregate (>9.5 mm) in asphalt mixtures influences the complexity of the pavement texture and the magnitude of the surface fractal dimension. Consequently, investigating the relationship between the proportion of coarse aggregate and the skid resistance of pavements has emerged as a novel research area. Huang et al. [35] studied the relationship between the concave–convex distribution characteristics of the pavement texture and the gradation, and Yun et al. [36] constructed a relationship function between the two, and initially determined the law between the initial skid resistance performance of the pavement and the gradation of the mixture, but the experimental results need more outdoor tests to verify the stability of the functional equation and should be further studied to identify the link between the type of pavement gradation and the law of the attenuation of the skid resistance performance. Compared with previous studies, Hou et al. [37] carried out research on the connection between 1D-textured, 2D-textured, and gradation-type surfaces, respectively, but the established model between texture and gradation had a limited range of applicability and multiple limitations, and the streamlined design of the model limited the subsequent development and extension of the model to a certain extent. The grading type of the mix and the pavement texture have different purposes in the characterisation of the skid resistance of asphalt pavements, and in general, identifying the relationship between the grading type and the pavement macro- and microtexture can help the construction of a comprehensive skid resistance grading design model.

As the macro-/microtexture of asphalt pavement shows a good positive correlation with the void ratio [38], pavement with a good macrotexture can improve the safety of vehicles travelling at high speeds as well as the anti-skid performance of asphalt pavement during rainy and snowy weather [39,40]. Yu et al. [41] conducted accelerated loading tests indoors on three gradation types, namely AC-13C, SMA-13, and GAC-13C, and analysed them using the empirical modal decomposition model (EMD) with the Hilbert–Huang transform (HHT). The results showed that the skidding resistance of the SMA-13 asphalt mixtures had the lowest decay rate. The macroscopic tectonic depth and pendulum value after 1 million axial loads were the largest in the SMA-13 mixes, which indicates that they have good skid resistance. Zhang et al. [42] investigated the skid resistance of three types of asphalt mixture pavements, namely Open-Graded Friction Course (OGFC), Stone Mastic Asphalt (SMA), and Antiskid Asphalt Mixture (AK-16A), and concluded that OGFC pavements have the best skid resistance, but the fatigue resistance of OGFC pavements is poor, and AK-16 asphalt pavements are average across all performance categories. The findings of Li et al. [43] also support this perspective. The SMA-16 asphalt mixture pavement exhibits satisfactory skid resistance during its early service period. The results of the research conducted by Liu et al. [44] on intermittent-graded asphalt mixtures for road performance demonstrate that these mixtures exhibit excellent resistance to rutting at high temperatures, excellent water stability, and superior durability. Moreover, they exhibited optimal macroscopic texture structure depth. Consequently, intermittent-graded asphalt mixtures are recommended as a superior surface material for roads [45,46,47]. In practice, the asphalt mixture of interrupted gradation pavement construction is susceptible to segregation. To address this issue, Li Zhi et al. [48] developed a novel gradation based on the CAFV method and interrupted gradation design theory [40,49].

2.2.3. Influence of Asphalt Binder on the Skid Resistance of Pavements

Asphalt binder plays a role as a binder material with properties like adhesion, thermal insulation, and waterproofing. The asphalt binder and fine aggregates, in conjunction with the mineral powder, combine to form asphalt slurry, which serves to enhance the interfacial adhesion between the aggregates. Asphalt mixtures can be classified into two distinct categories based on the manufacturing process and the temperature at which it is utilised. These categories are hot mix asphalt (HMA) and warm mix asphalt (WMA), respectively. The Salient details and frictional properties of asphalt mix classifications are shown in

Table 1. Salient details and frictional properties of asphalt mix classifications.

Asphalt Mix Type	Texture-Based Mix Category	Classification Details and Friction Characteristics [31]
Hot Mix Asphalt	Continuous-graded asphalt mixture	Continuous-graded asphalt mixtures are made of continuous-graded, mutually embedded dense aggregates and asphalt in a hot mixing state; these mixtures form a hot paving state, and their macrotexture depths are generally 0.4–1.2 mm, but their microtextures are similar to open-graded and semi-open-graded asphalt mixtures [50].
	Intermittent-graded asphalt mixture	Intermittent-graded asphalt mixtures consist of aggregates that lack one or more grades in the gradation composition and asphalt in the hot mixing and hot paving forming mix; their macrotexture depths are usually greater than 1 mm, and coarse aggregates creating a macrotexture are best, as they provide friction in wet weather and low tyre noise [51].
	Open-graded asphalt mixture	Open-graded asphalt mixtures consist mainly of coarse aggregates, with fewer fine aggregates, and macroscopic texture depth ranges from 1.5 to 3 mm for mixes with residual voids greater than 15 percent after compaction, but the microscopic texture remains similar to that of other types of mixes [52].
Warm Mix Asphalt		Warm mix asphalt consists of aggregates mixed with asphalt at room temperature to form a mix that is environmentally friendly, with less dust and fumes, and exhibits very few differences between its macrotexture and microtexture compared to HMA pavements, and no significant differences in early skid resistance have been observed [53].

.

The skid resistance of asphalt pavements is primarily influenced by the characteristics of the materials used, including the aggregate, mix gradation, and asphalt binder. Additionally, the skid resistance can be affected by vehicular factors, such as the tyre tread design, the contact area between the tyre and road, tyre inflation pressure, and the speed at which the vehicle travels, as well as environmental conditions.

3. Methods for the Characterisation and Evaluation of the Skid Resistance of Asphalt Pavements

3.1. Characterisation Methods for Asphalt Pavement Macro- and Microtextures

PIARC classified asphalt pavement textures into four categories based on wavelength and amplitude, as shown in

Table 2. Texture structure classification and formation reasons.

Texture Type	Wavelength (mm)	Amplitude (mm)
Unevenness	>500	1~200
Large texture	50~500	1~50
Macrotexture	0.5~50	0.2~10
Microtexture	<0.5	0~0.2

. In practical engineering applications, the key factors affecting the slip resistance of pavements are macrotexture and macrotexture (Figure 4). In this context, microtexture is defined as being within the range of λ (wavelength) < 0.5 mm and A (amplitude) < 0.5 mm. The microtexture of pavements is primarily influenced by factors such as the shape, angularity, and textural properties of the aggregate particles [54,55]. Macrotexture is defined as being around 50 mm in size, with a wavelength of between 0.5 mm and 50 mm and an amplitude of between 0.1 mm and 20 mm. The macrotexture observed in a pavement primarily depends on several factors, including the gradation, shape, and particle size of the mix [56]. An oversized texture in pavements is defined as ranging from 50 mm < λ < 500 mm to 0.1 mm < A < 50 mm, and the formation of the oversized texture mainly depends on pavement material, construction quality, construction technology, service environment, vehicle axle load, and other factors. Nevertheless, related studies have demonstrated that pavement texture can also be characterised by surface roughness, surface waviness, and shape error. Of these characteristics, surface roughness refers to the microgeometric shape of an object’s surface, which is characterised by small spacings and peaks and valleys. This also belongs to the category of surface micro-shape. Surface roughness is different from surface macrotexture (e.g., straightness, roundness, etc.) and surface waviness. In the field of mechanical and friction research, surface roughness is usually divided by the size of the ratio of wave pitch (the distance between two neighbouring peaks or troughs) and wave height; a ratio of more than 1000 is a macrotexture, a ratio of less than 50 is a microtexture, and a ratio between 1000 and 50 is known as waviness. Surface roughness can also be divided according to the size of the wave spacing—wave spacing of more than 10 mm is considered part of a macrotexture; a wave distance of less than 1 mm is considered part of a microtexture; and a wave distance between the above two or with periodic changes is considered part of a wave degree. The above two types of road surface texture are only characterised differently, there are no fundamental differences. By analysing the way in which multiple pavement textures are characterised, the following conclusions can be drawn: microtexture predominantly affects skid resistance under low-speed and dry pavement conditions, whereas macrotexture is the primary contributor to frictional performance at high speeds on wet pavements [57]. The presence of macrotexture enables effective surface water drainage through the formation of flow channels, thereby minimising the likelihood of hydroplaning. Conversely, microtexture improves tyre–pavement adhesion by increasing the actual contact area between the tyre and the pavement surface. This effect becomes particularly significant under wet conditions where a thin water film exists, thereby enhancing skid resistance in such scenarios.

3.2. Pavement Macro- and Microtexture Measurement Methods

3.2.1. Pavement Texture Measurement Methods

(1)

Laser scanning measurement method

The laser texture scanning method employs a laser transmitter to emit a laser beam onto the pavement surface and to receive the reflected laser beam from another angle. This enables the construction depth of the pavement to be accurately and rapidly determined. However, the equipment required for this method is expensive and it is not possible to measure the pavement when it is submerged in water [58]. The most commonly used laser texture scanning instruments are circular texture meters (CTMs) [59] and laser texture scanners (LTSs), as shown in Figure 5. A CTM is a fixed-point, non-contact laser texture testing device. The CTM measures the road surface profile at 0.868 mm intervals along a circular path with a diameter of 286 mm. The CTM device rotates at a speed of 0.1 m-s-1 to generate 8 segments of contour lines of the road surface and stores the data in a portable computer. Based on these contour lines, two different pavement macrotexture indices can be calculated: the mean section depth and the root mean square of the contour [60,61]. The LTS is capable of measuring the macrotexture and microtexture of pavements in a straightforward manner and with an excellent resolution.

(2)

Image processing methods

The image processing method is based on the principle of diffuse reflection. Intelligent devices such as high-definition cameras are employed to obtain the three-dimensional parameters of the structure of the road surface. The digital images are then converted into grey-scale images by computer algorithms. The combination of pavement textures was quantitatively analysed by image processing methods [62]. The efficacy of image processing methods is contingent upon a multitude of variables, including lighting conditions, road conditions, camera positions, and other factors [63]. Despite this, image processing methods are identified by their accuracy, the quantity of information they yield, and their ease of operation, as shown in Figure 6.

(3)

CT scanning method

The CT scanning method utilises industrial CT to obtain three-dimensional morphology images of asphalt mixtures. This is achieved through the use of software to screen, distinguishing between coarse and fine aggregates, asphalt binder, and other substances formed by the texture condition [64]. The CT equipment is illustrated in Figure 7. The CT scanning method is more accurate but is limited to indoor measurements with inefficient detection [65].

(4)

Mechanical stylus method

The mechanical stylus method works by moving the tip of a stylus over the road surface at a certain speed, and the computer uses sensors to detect the undulation signals of the road surface, which contain the macroscopic texture of the road surface; the working principle is shown in Figure 8. The mechanical stylus method offers several advantages, including high data accuracy, reliable and stable measurement outputs, excellent repeatability, minimal sensitivity to environmental conditions, and straightforward instrument operation. However, this technique also presents certain limitations: it is susceptible to causing surface scratches on the measured object, exhibits relatively low scanning speeds, and is constrained by a limited measurable area [66]. The measurement accuracy of the mechanical stylus method is 0.02 to 5 μm. The detection error rate of the mechanical stylus method is related to the size and hardness of the stylus, the signal acquisition technology, and the speed of movement.

(5)

Sand patch method (SPM)

The sand patch method (SPM) evaluates pavement macrotexture by distributing a known volume of sand onto the surface and measuring the diameter of the sand-covered area. The mean texture depth (MTD) is subsequently determined by dividing the sand volume by the area of the resulting circular patch [67]. SPM can be implemented using either a manual or an automated (electric) approach. The electric version replaces manual operations with mechanised procedures, thereby minimising the influence of operator variability and improving measurement consistency [68]. The corresponding calculation of MTD is presented in Equation (4), and the schematic of the SPM setup is illustrated in Figure 9.

$${\mathrm{MTD}=4\mathrm{V}/(\pi \mathrm{D}}^2)$$

(4)

where MTD is the mean texture depth of pavement, V is the volume of the sand, and D refers to the average diameter.

3.2.2. Measurement of Pavement Microtexture

(1)

Laser measurement method

At present, pavement microtexture is predominantly assessed using non-contact optical techniques. The most commonly employed instruments include circular texture meters (CTMs), linear laser scanners (LLSs), and laser texture scanners (LTSs) [59,69,70]. CTMs capture surface texture by following a circular scanning trajectory, whereas the LLSs operate along a linear path. A LTS is capable of simultaneously acquiring both the macrotexture and microtexture characteristics of a pavement surface.

(2)

Direct microscopic observation method

The direct microscopic observation method enables the evaluation of the aggregate surface microtexture by carrying out a high-resolution visual inspection of surface features at the microscale [71], in which SEM can magnify the original image of the aggregate surface by a factor of 6000 [72], the texture condition can be directly binarised by software such as Fractalfox, and finally, the microtexture of the aggregate surface can be analysed [73].

In addition, atomic force microscopy (AFM) has shown great potential for application during microtexture studies of pavements [68]. A number of researchers have conducted computer simulations of the molecular dynamics of microfriction between a tyre and an asphalt pavement. The atomic force microscope (AFM) was found to be an effective tool for validating simulated microtexture changes in the pavement. The AFM is shown in Figure 10.

3.3. Traditional Skid Resistance Testing Methods and Indicators for Asphalt Pavements

Traditional test indexes for pavement skid resistance are as follows: friction coefficient of pavement, dynamic friction coefficient, transverse friction coefficient, longitudinal friction coefficient, and slip index.

The coefficient of friction of a road surface can be quantified by a British pendulum tester, which utilises the free-fall motion of a pendulum at a fixed height to convert the potential energy of the pendulum into kinetic energy. This is achieved through the use of a rubber pad on the pendulum which comes into contact with the road surface at a slip speed of 10 km/h [75], thereby determining the characterised friction coefficient of the road surface under wet conditions. The friction coefficient of pavement can be employed to assess the skid resistance characteristics of a road surface when a vehicle is in motion. BPN is associated with the aggregate type, shape, and particle size. The method is inexpensive and requires minimal equipment and operation, making it a popular choice for laboratory and engineering tests. Nevertheless, the test outcomes are susceptible to variability resulting from operator influence, the thickness of the water film, and the ageing condition of the rubber block [76]. A British pendulum tester is depicted in Figure 11.

Dynamic friction testers (DFTs) are used to determine the dynamic coefficient of friction between two sliding contact surfaces. In the Field Test Methods of Highway Subgrade and Pavement (JTG 3450-2019) [77] implemented by the Ministry of Communications of China on 1 September 2008, DFT was identified as one of the test methods for determining the coefficient of friction of pavement. The coefficient of friction, as measured by the DFT at a speed of 60 km/h, is employed as the standard measurement, designated as DFT₆₀. Nevertheless, the test results may be inaccurate in the presence of pollutants on the road. The test setup is depicted in Figure 12.

Pavement skid resistance can be further subdivided into longitudinal and transverse skid resistance and is characterised by the transverse and longitudinal friction coefficients, respectively.

The transverse force coefficient is the ratio of the transverse force to the vertical force. The transverse force coefficient SFC60 was quantified by the lateral force coefficient test vehicle at a velocity of 60 km/h while the road surface was kept in a wet state, as illustrated in Figure 13. However, due to the high cost of the test equipment and its inability to test small areas and other specialised pavements [78], the transverse force coefficient test is somewhat limited in its use and dissemination.

The longitudinal friction coefficient determines the emergency braking distance of a vehicle and is closely linked to driving safety. Li et al. [79] used the pavement longitudinal coefficient of the friction testing vehicle. Since the longitudinal friction coefficient test vehicle must be on a road section of specific length (generally ≥ 100 m) to obtain data, the test range needed to be determined in advance. If the test section was only laid out temporarily at the site to carry out the test, the test cost would be too high. The device is shown in Figure 14.

The slip resistance of the road surface can also be expressed using the slip index (SN). The test equipment is a locked wheel tester with the test wheel which can be fully locked. The tyres were inflated to a pressure of 165 kPa and were of the standard tread type. In the test project, the water film thickness was 0.5 mm, which equates to the distance between the test tyre and the pavement being 0.5 mm. The test utilised the slip index (SN) to identify the skid resistance of the pavement, as demonstrated in Equation (5). The locked wheel tester was able to test the coefficient of friction of pavement continuously, but the test error could have been affected by the sprinkler system [80].

$$\mathrm{SN}={\displaystyle \frac{\mathrm{F}}{\mathrm{W}}}\times 100\%$$

(5)

where F is the friction force acting on the test wheel, N. W is the vertical load acting on the wheel, N.

3.4. Asphalt Pavement Skid Resistance Evaluation Model

Traditional pavement macro- and microtexture tests are conducted under controlled conditions; however, the type of equipment, test speed, test temperature, and humidity all influence the test results during actual pavement skid resistance testing [81,82]. To accurately evaluate pavement skid resistance, researchers have established various models to describe the variability in skid resistance by integrating the effects of environmental factors on test results [83,84]. The most commonly used models include the Empirical Formula Model, the Penn State Model, the Modified Penn State Model, the Asymptotic Model, the PIARC Model, the International Friction Index Model [22], the Pavement Skid Resistance Prediction Model, and the Fractal Difference Function Model, as shown in

Table 3. Evaluation models for pavement skid resistance.

Model Type	Parametric Expression
Empirical Formula Model	$\left(1\right) \mathrm{Y}=\mathrm{AInx}+\mathrm{B}$ $\left(2\right) \mathrm{Y}=\mathrm{A} \mathrm{e} \mathrm{B} \mathrm{x}$
	Y is the slip resistance index or the depth of construction, x is the number of axle loads (10,000), and A and B are regression coefficients.
Penn State Model	${\mathrm{F}\left(\mathrm{S}\right)=\mathrm{F}}_0\exp \left(1-{\displaystyle \frac{\mathrm{S}}{{\mathrm{S}}_0}}\right)$
	F(S) represents the friction factor of the wheel at a slip speed of S; F0, S0 are the characteristic parameters of the test apparatus; F0 is the friction factor at a slip velocity of 0; and S0 is a function that depends on the macroscopic texture of the pavement.
Modified Penn State Model	${\mathrm{F}\left(\mathrm{S}\right)=\mathrm{F}}_{10}\exp ({\displaystyle \frac{10-\mathrm{S}}{{\mathrm{S}}_0}})$
	F10 is the friction factor for a slip velocity of 10 km·h−1, and the rest of the parameters have the same meaning as in the Penn State Model.
Asymptotic Model	${\mathrm{y}=\mathrm{Ae}}^{{\mathrm{B}}_{\mathrm{x}}}+\mathrm{C}$
	A is the attenuation amplitude, B is the attenuation rate, C is the final attenuation value, and x is the number of axial loads.
PIARC Model	${\mathrm{F}}_{60}{=\mathrm{A}}^{\prime }{+\mathrm{B}}^{\prime }{\mu }_{\mathrm{s}}^{\prime }\exp [{\displaystyle \frac{\mathrm{S}-60}{{\mathrm{a}+\mathrm{bT}}_{\mathrm{x}}}}{]+\mathrm{C}}^{\prime }{\mathrm{T}}_{\mathrm{x}}$ ${\mathrm{S}}_{\mathrm{p}}{=\mathrm{a}+\mathrm{bT}}_{\mathrm{x}}$ ${\mathrm{I}}_{\mathrm{IFI}}{(\mathrm{F}}_{60}{,\mathrm{S}}_{\mathrm{p}}{)=\mathrm{F}}_{60}\exp ({\displaystyle \frac{\mathrm{S}-60}{{\mathrm{S}}_{\mathrm{p}}}})$
	$S_{\mathrm{p}}$ is the speed number; $F_{60}$ is the friction factor corresponding to a slip speed of 60 km·h−1; $A^{\prime },B^{\prime },C^{\prime }$ are the influence coefficients; $T_{\mathrm{x}}$ is the pavement construction parameter (MTD when measured by the sand-laying method, and MPD when measured by the laser method); $a,b$ are the calibration parameters of the instrument$; {\mu }^{\prime }$ is the friction factor at a slip speed of the s friction factor when the slip velocity is s; and IFIis the international friction index.
Pavement Skid Resistance Prediction Model	$\mathrm{y}=-\mathrm{Aln}\left(\mathrm{x}\right)+\mathrm{C},\mathrm{A}=0.178\mathrm{s}+1.9054$
	y is the pendulum value; A is the attenuation rate; X is the cumulative number of axle loads, i.e., 10,000 times; C is the initial value of attenuation related to asphalt mix gradation; and s is the aggregate polishing value.
Friction and Texture Models	${\mathrm{BPN}=\alpha +\beta }_{{\mathrm{MPD}}_{\mathrm{maxo}}}{\mathrm{Macro}}_{\mathrm{MPD}}{+\beta }_{\mathrm{treat}}{\mathrm{Treat}+\beta }_{\mathrm{treat}}{\mathrm{Micro}}_{{\mathrm{param}}_{\mathrm{i}}}^{{\mathrm{bascline}}_{\mathrm{j}}}$
	${\mathrm{Macro}}_{\mathrm{MPD}}$ is the MPD; Treat is a virtual variable that takes 1 when the cross-section receives an optical texture and 0 otherwise; and ${\mathrm{Micro}}_{{\mathrm{param}}_i}^{{\mathrm{bascline}}_j}$) is the microtexture baseline space parameter.

.

The Empirical Formula Model is as follows: $\left(1\right) \mathrm{Y}=\mathrm{AInx}+\mathrm{B}$ $\left(2\right) \mathrm{Y}=\mathrm{A} \mathrm{e} \mathrm{B} \mathrm{x}$. Chen et al. [85] analysed the change rule of pavement skid resistance using these two models. The results demonstrated that these two models were unable to simulate the process of pavement skid resistance attenuation caused by the pre-compaction stage. Nevertheless, the logarithmic curve model is capable of accurately representing the stability stage in the later phase of asphalt pavement skid resistance decay.

The Penn State Model evaluates the skid resistance of pavements by investigating the relationship between the pavement friction factor F and the test wheel slip velocity S. In this case, the slip speed S is directly related to the macroscopic texture of the pavement, and S can be classified into three cases according to the test equipment: ① Locking wheel friction coefficient metre: S = V; V is the speed of the test vehicle. The same as below. ② Fixed slip rate friction coefficient metre: S = V − K; K is the slip rate. ③ Lateral force type friction coefficient metre: S = V − sinα; α is the deflection angle of the deflection wheel. Due to the low test speed of the friction factor measuring instrument at this stage and despite the fact that the speed during the test is not 0, the friction factor of the test equipment at low speeds can only be obtained by calculation. The smaller values of S₀ indicate that the friction factor of the road surface is high at lower slip velocities. Conversely, the larger values of S₀ imply that the velocity has a small effect on the friction factor [86]. Fan et al. [87] used the friction factor calculated at a slip velocity of 10 km/h instead of the friction factor F₀ at a slip velocity of 0 and F₁₀ = 10 km/h to obtain the modified Penn State model.

The modified Penn State Model is only capable of evaluating the relationship between friction factor and speed at a speed of 10 km/h. This model is not applicable to the full range of test equipment due to the instrument’s test speed varying widely and being unable to be controlled stably at 10 km/h.

The Asymptotic Model is able to predict the asphalt pavement pendulum value and the construction depth, respectively. When x = 0, y = A + C represents the initial skid resistance value of the asphalt pavement. When x→∞, y = C denotes the final value of the pavement after the decay of skid resistance [88]. Shuang et al. [89,90] conducted accelerated loading tests on AC-13, SMA-13, and OGFC-13 mix pavements using the Asymptotic Model. They concluded that the predicted results of the Asymptotic Model were closely related to the number of vehicle axle loads.

The PIARC model was developed by the World Road Association following a comprehensive study of 54 road sections in Belgium and Spain, utilising 47 types of pavement macro- and microtexture testing equipment. The PIARC model incorporates all test variables and enables the conversion of measured pavement friction coefficients at various test parameters to the friction coefficient at 60 km/h [69,91]. Where the International Friction Index (IFI) consists of two parameters, i.e., speed Sp and the friction factor F₆₀, at the standard speed, where ${\mathrm{S}}_{\mathrm{p}}{=\mathrm{a}+\mathrm{bT}}_{\mathrm{x}}$; a and b are regression coefficients; and T_x is the depth of the pavement construction. The skid resistance of most pavements can be evaluated by the PIARC model, as shown in Figure 15. It has been divided into four distinct zones in Figure 15. Zone I exhibits excellent skid resistance. In Zone II, Sp is relatively small, indicating that the macroscopic texture of the pavement requires improvement. In Zone III, F60 is relatively small, indicating that the pavement microtexture requires improvement. In Zone IV, Sp and F60 are both small, indicating that both the macrotexture and microtexture of the pavement require improvement [92]. Currently, the majority of countries utilise the International Friction Index (IFI) as the standardised evaluation index for pavement skid resistance.

A pavement skid resistance prediction model was investigated using image processing techniques to analyse the texture of mixtures with different aggregates and gradations [93]. This model can accurately evaluate the skid resistance of asphalt pavements based on the characteristics of the aggregates and gradations [94,95]. The pavement skid resistance prediction model focuses on the influence of test parameters, including aggregate abrasive properties, mixture gradation, and traffic volume, on the prediction of asphalt pavement skid resistance. This has a wide range of applications and can be used as a reference for the evaluation of asphalt pavement skid resistance [96]. Nevertheless, the model does not consider the morphological characteristics of aggregates, and thus the accuracy of the regression model needs to be improved.

The friction versus texture model, developed by Kane et al. [97] through multiple regression analyses, describes the relationship between pavement skid resistance and both macrotexture and microtexture, as shown in Table 3. The friction and texture model predicts the slip resistance of pavements through the use of pavement texture parameters, as illustrated in Figure 16 [98,99].

4. Study of the Relationship Between Pavement Macroscopic and Microscopic Texture Characteristics and Skid Resistance Performance

The macroscopic texture feature parameters of the pavement affect the anti-slip performance of the pavement, and the contributions of macroscopic texture feature parameters, microtexture feature parameters, and travelling speed to the anti-slip performance of the pavement is shown in Figure 17. At low vehicle speeds, the microscopic texture of dry and wet pavement exerts the greatest interaction force on tyres [100]. Under higher vehicle speed conditions, the macroscopic texture of the tyre’s tread with the wet road surface facilitates the drainage of water, and then the interaction force between the microscopic texture and the tyre becomes the dominant friction force [101].

4.1. Effect of Macrotexture on Skid Resistance of Asphalt Pavements

4.1.1. Dry Pavement

The size and number of pavement microconvex bodies are the primary determinants of the macroscopic texture of the pavement. However, the contact process between the pavement microconvex bodies and the tyre, the vehicle’s travelling speed, and the average spacing between the pavement microconvex points must be satisfied in order to fulfil Equations (6)–(8). The angular velocities of the tyre deformation (ω) and the travelling speeds in the travelling process must also satisfy Equation (6). When the deformable surface of the tyre comes into contact with the microconvex bodies, the accumulated dissipated energy of the tyre increases with the height of the microconvex bodies [103], and the accumulated dissipated energy of the tyre tends to increase with the decrease in the average spacing of the microconvex bodies of the road surface [104]. This is due to the fact that a reduction in the spacing of the pavement microconvexes also increases the contact area between the tyre and the road surface, thus increasing the friction between the tyre and the road surface [105].

$$\omega =2\pi \nu /\mathrm{D}$$

(6)

$${\mathrm{E}}_{\mathrm{c}}{=n\mathrm{E}}_{\mathrm{loss}}={\displaystyle \frac{\mathrm{l}\omega }{2\nu }}{\mathrm{E}}^{\prime }{\sigma }_{\mathrm{m}}^2\sin \delta$$

(7)

$${\mathrm{E}}_{\mathrm{c}}{=n\mathrm{E}}_{\mathrm{loss}}={\displaystyle \frac{\pi \mathrm{l}}{\mathrm{D}}}{\mathrm{E}}^{\prime }{\sigma }_{\mathrm{m}}^2\sin \delta$$

(8)

where v is the linear velocity, E_c is the cumulative dissipated energy, and E_loss is the dissipated energy.

4.1.2. Wet Pavement

In 2013, Najafi et al. proposed the concept of three skid-resistant regions for a pavement. This concept explains the interaction between the macroscopic texture of the pavement and the tyre-contact region on a wet pavement [106]. In Zone 1, the water flows through the macroscopic texture of the road surface, increasing the contact area between the tyre and the road surface and reducing the effect of the water’s lubricating action on the contact position. In Zone 2, the pavement microtexture plays a crucial role in reducing the thickness of the surface water film. Based on Bernoulli’s energy principle, the dynamic water pressure generated by a tyre on a wet pavement surface increases quadratically with vehicle speed [107]. As vehicle speed increases, the corresponding rise in dynamic pressure promotes a progressive thinning of the water film layer [108]. In Zone 3, direct contact between the tyre and the pavement surface is established. The main mechanisms contributing to frictional interaction are adhesion and hysteresis. During this phase, pavement skid resistance is significantly affected by environmental factors, the viscoelastic properties of the tyre rubber, and the macrotexture of the pavement surface. Under rainfall conditions, a lubrication effect develops, which results in a substantial decrease in adhesive forces and a corresponding modification of the hysteresis component [109]. The hysteresis force depends mainly on the structural characteristics of the pavement, but for the same maximum nominal aggregate size, an appropriate increase in the macroscopic texture depth of the pavement will improve the wet pavement skid resistance.

4.2. Effect of Microtexture on Skid Resistance of Asphalt Pavements

4.2.1. Dry Pavement

When vehicles are travelling at low speeds (travelling speeds below 30–50 km/h), the microtexture of the asphalt pavement plays a decisive role in the anti-slip properties, which mainly depend on the roughness of the aggregate surface [110]. The influence of microtexture characteristic parameters on the skid resistance of pavements depends on three main points: height parameter, density, and sharpness. Good microtexture feature parameters can increase the contact area between the tyre and the road surface and improve the skid resistance of the road surface. As such, the microtexture of the pavement ensures the safety of driving cars in bad weather. The microconvex bodies of smaller sizes on the pavement will form microcutting effects on the tread under driving loads, and the resistance generated during microcutting constitutes part of the friction between the tyre–road surface [111]. When the road microtexture touches the tyre surface, the interaction force between the two is generated by the energy loss from tyre adhesion and microcutting, and the adhesion friction can be derived from Equation (9).

$${\mathrm{F}}_{\mathrm{A}}{=\mathrm{K}}_1{\mathrm{K}}_2{\sigma }_{\mathrm{m}}{\displaystyle \frac{\mathrm{N}}{\mathrm{H}}}\tan \delta$$

(9)

where ${\mathrm{K}}_1$ is the adhesion coefficient, ${\mathrm{K}}_2$ is the microcutting coefficient, ${\sigma }_{\mathrm{m}}$ is the maximum stress at the tip of the microconvex body, N is the vertical load, H is the hardness of the rubber tyre, and δ is the hysteresis angle. H and δ are determined by the properties of the tyre material. Accordingly, the greater the height parameter and sharpness of the microtexture of the road surface, the greater the depth of the tyre embedded in the microconvex bodies. The greater the density of the microtexture, the greater the number of points of contact between the road surface and the tyre. This results in a larger σ_m and an increase in the friction between the tyre and the road surface [112]. Consequently, an increase in the height, density, and sharpness of the microtexture of the pavement is beneficial for improving the skid resistance of the pavement.

4.2.2. Wet Pavement

For wet road surfaces, using the same principle as that utilised in Section 4.2.1, the microtexture also plays a key role in the skid resistance of low-speed vehicles. The moderate height and sharpness of the microconvex body can enhance the safety of driving in wet conditions. When there is water on the pavement of a road, the microconvex bodies directly pierce the water film and are embedded in the rubber tyre, which is conducive to creating direct contact between the road surface and the rubber tyre, increasing the contact area between the tyre and the road, and preventing the vehicle from skidding [113]. The density of the microtexture determines the effective contact surface between the tyre and the road pavement; the higher the density of the microtexture, the more points of attachment of the microconvex bodies to the tyre per unit area, and a large number of concentrated stresses are generated at the locations where the road surface is in contact with the tyre [114]. Therefore, the height characteristics, density, and sharpness of the microtexture can help to improve the anti-skid performance of wet road pavement surfaces [9,91,115].

5. Conclusions and Future Work

From the perspective of the relationship between pavement macroscopic texture and skid resistance, this article reviewed the current research status and development trends in related fields both domestically and internationally. In the study of the anti-skid mechanism of asphalt pavements, the interaction forces between tyres and pavement, as well as the factors affecting the anti-skid performance of asphalt pavements, were summarised. Regarding the characterisation and evaluation methods of asphalt pavement skid resistance, various methods for characterising the macro- and microtextures of asphalt pavements were reviewed. In research on the relationship between pavement macro- and microtextures and skid resistance, the effects of pavement texture on skid resistance have been investigated on both dry and wet pavements. Although numerous useful research findings have contributed to improving road conditions and safety, several shortcomings remain.

• The skid resistance of asphalt pavements is significantly influenced by the type and quality of the aggregate in the asphalt mix and the grading type of the asphalt mix. While different HMA classifications often exhibit similar texture characteristics, the relationships between aggregate properties and grading design and the endogenous mechanisms of macroscopic and microscopic texture refinement for mix differentiation have not been thoroughly investigated.
• In terms of pavement skid resistance measurement methods, there remains a significant gap between laboratory testing and field testing. It is necessary to optimise pavement skid resistance testing technology to bridge this gap between laboratory and actual engineering tests and to establish a quantitative relationship between pavement texture characteristics and skid resistance.
• Based on the various random uncertainty parameters in actual engineering, the macro- and microtexture evolution behaviour of pavement under the service conditions of the entire environmental domain has not yet been fully elucidated. It is necessary to investigate the long-term skid resistance attenuation characteristics of different pavement types throughout their life cycles and to integrate these uncertainty factors with pavement skid resistance to develop a unified real-time skid resistance evaluation model for asphalt pavements.