Impact of Aggregate Characteristics on Frictional Performance of Asphalt-Based High Friction Surface Treatments

Abstract

High Friction Surface Treatments (HFST) are recognized for their effectiveness in enhancing skid resistance and reducing road accidents. While Epoxy-based HFSTs are widely applied, they present limitations such as compatibility issues with existing pavements, high installation and removal costs, and durability concerns tied to substrate quality. As an alternative to traditional Epoxy-based HFSTs, this study investigated the effects of aggregate gradation as designated by agencies on the performance of asphalt-based HFST. Various aggregate types were assessed to evaluate friction performance and the impact of polishing cycles on non-Epoxy HFST. It was found that adjustments in aggregate size and gradation may be necessary when transitioning to asphalt-based HFSTs, given the different nature of asphalt as more temperature susceptible compared to Epoxy. Various asphalt binder grades were considered in this study. A series of tests, including the British Pendulum Test (BPT), Dynamic Friction Tester (DFT), Circular Track Meter (CTM), Micro-Deval (MD), and Aggregate Imaging Measurement System (AIMS), were conducted to measure Coefficient of Friction (COF), Mean Profile Depth (MPD), texture, and angularity before and after polishing cycles. The results showed that the COF in asphalt-based slabs decreased more significantly than in Epoxy-based slabs as polishing cycles increased for HFST and medium gradations. However, in coarse gradation, the COF of slabs using asphalt-based binder matched or even surpassed that of Epoxy after polishing. Notably, the PG88-16 binder for Calcined Bauxite (CB) had the smallest reduction in COF after 140K polishing cycles, with only a 19% decrease compared to a 23% reduction for Epoxy.

1. Introduction

Traffic crashes, injuries, and fatalities are costly and continue to be one of the major challenges faced by state DOTs and local highway agencies. As recognized, road safety encompasses more than just promoting safe driving; pavement surface maintenance plays a vital role in preventing traffic accidents or reducing their severity [1].

High Friction Surface Treatment (HFST) is proving to be a very effective and cost-efficient method of treatment in improving pavement friction and reducing collisions dramatically in a wide array of traffic conditions around the world. In the past 15 years, the implementation of HFST has significantly increased throughout the United States; most states are now applying this technology to enhance traffic safety in various locations. The wide acceptance of HFST is proof of its importance for accident reduction as a strategic measure. This treatment is very helpful, especially at sharp curves, intersections, and steep grades where high levels of pavement friction are imperative in nature to avoid vehicle skidding and loss of control. In addition to saving lives, HFST reduces crash rates and reduces the economic traffic crash costs by greatly enhancing surface traction [2]. HFST is typically composed of a thin layer of polymer binder with an aggregate resistance to polishing. This combination provides an improved friction characteristic of the pavement surface by reducing skidding and accidents and presents a long-lasting solution to increase the safety of the roads [3]. HFST with Epoxy resins conforming to AASHTO MP 41 is normally used in the United States. This specification ensures that HFST binder performs well and has a durable nature. AASHTO MP 41 specification requires the use of refractory-grade Calcined Bauxite (CB) with at least 85% aluminum oxide [4]. This high-grade material was selected due to its exceptional hardness and its resistance to polishing, which were necessary to provide long-term pavement friction, thereby reducing accidents.

In many cases, HFST may be subject to a variety of integrity problems and early degradation such as aggregate loss, cracking, debonding, and delamination. Early distresses in HFST can be caused by a multitude of factors. The condition of the current pavement as well as the Epoxy resin binder. In HFST, the resin binder interlocks with the pavement surface through macrotexture, a portion of the mechanical bond between the binder and pavement surface. A weak mechanical bond can be formed when the pavement surface lacks adequate macrotexture. Another major cause of low bond strength is poor preparation of the surface, where the surface should have been cleaned but still has dust and debris; this interferes with the bond between the resin binder and pavement surface. Such circumstances might lead to the delamination of the HFST from the pavement surface. It is such a situation that could have caused this failure of HFST [5,6]. In addition to its higher stiffness, the coefficient of thermal expansion (CTE) of Epoxy-based HFST has been considerably higher than that of hot-mix asphalt (HMA). Due to this inconsistent CTE, the HFST and old pavement suffer from thermal incompatibilities, thus generating differential thermal displacements of the HFST from existing pavement under temperature changes. The resulting temperature-induced differential motion may induce additional distress mechanisms into the HFST system, thus its ultimate failure [7].

In response to these concerns, the New Jersey Department of Transportation created a treatment called High Friction Chip Seal (HFCS) to address compatibility issues between HFST and existing asphalt pavements. This approach uses an asphalt binder instead of typical Epoxy-based HFST, which is more compatible with the pavement substrate. The Missouri Department of Transportation (MoDOT) also performed studies to assess the costs and efficacy of surface treatments created with asphalt-based binders [8,9].

Recently, state agencies have increasingly focused on developing asphalt-based alternative binders for HFST applications. An example is the National Cooperative Highway Research Program (NCHRP) RFP #NCHRP 10-145, which emphasizes the importance of optimizing aggregate selection for HFST [10].

This paper presents the investigation of various aggregates with different gradations and develops some asphalt-based alternatives to the conventional Epoxy resin binder in HFST applications. The present study focused on the determination of optimum aggregate gradation for asphalt-based HFST, taking into account the differences existing between asphalts with respect to Epoxy resin binders regarding mechanical characteristics such as stiffness, bonding mechanisms, and other differentiating factors.

Adjusting the aggregate gradation and type relative to standard gradations for HFST is essential to fully understand these differences and enhance durability and performance. The research dealt with finding a more efficient and appropriate HFST solution for current pavement substrates by investigating interactions of different types of aggregate and gradation with asphalt binders, thereby possibly providing a much more effective and sustainable solution toward keeping high pavement friction and long-term road safety.

2. Materials and Methods

2.1. Aggregate

Material properties of HFST aggregates are studied in this research with four different types, each having three different gradations. Calcined Bauxite (CB), Rhyolite (Rhy), Meramec, as well as Flint aggregates. Each aggregate was tested with three different gradations: a standard gradation that is often used in HFST applications in the US and referred to here as “HFST original size”; a coarser gradation labeled “coarse size”; and an intermediate gradation referred to as “medium size”. The different gradations were selected due to the inherent differences between Epoxy and asphalt binders and following specific standards by the sponsoring agency, the Missouri Department of Transportation (MoDOT). Epoxy’s stiffness supports fine particle bonding, while asphalt’s viscoelastic nature benefits from coarser gradations to improve bonding and aggregate retention [9].

Table 1. Aggregate Properties.

Properties	CB	Rhy	Meramec	Flint
Bulk specific gravity	3.25	2.56	2.45	2.51
Water absorption (%)	2.5	0.9	2.6	2.2
Uncompacted Void Content (UVC) (%)	44	42	41	49
LAA value (%)	Grade D	D	D	D
	16	17	15	19
MDA value (%)	15 min/30 min	15 min/30 min	15 min/30 min	15 min/30 min
	2.45/4.2	2.6/4.74	1.4/2.78	2.22/4.48

Note: LAA = Los Angeles abrasion, MDA = Micro-Deval abrasion.

lists the properties of these aggregates, and Figure 1 and Figure 2 present their gradations. CB, which is required by the AASHTO MP 41 specification as the standard aggregate for HFST, was chosen for its exceptional abrasive qualities, high hardness, and ability to maintain friction over time. Rhy, as noted by the Missouri Geological Survey, is a dense, fine-grained rock known for its resistance to weathering [11,12]. Meramec is commonly used in both road construction and concrete applications, while Flint is specifically produced to enhance friction in surface treatments such as chip seals [13].

2.2. Binder

In addition to Epoxy resin (FasTrac CE330, Lee’s Summit, MO, USA), a series of asphalt-based binders were considered as potential alternatives to Epoxy resin. Both performance grade and modified binder options were selected, which are known to be compatible with the selected aggregates and possess properties that optimize frictional characteristics as HFST binders. The PG binders used in this study included PG76-22, PG94-10, and a blend of these two binders. According to the supplier’s guidelines (Pure Asphalt Co., Chicago, IL, USA), PG94-10 is an air-blown asphalt binder with enhanced aging properties, setting it apart from conventional polymer-modified binder. Preliminary tests on trial samples using the British Wheel polishing device (Wessex Precision Instruments Ltd., Banwell, UK) showed that PG76-22 was too soft, while PG94-10, with its high stiffness, tended to lose aggregate. The ideal asphalt binder needs to withstand high temperatures to prevent aggregates from loosening in hot weather while also remaining durable in colder regions. As a result, blends of these two binders were tested, specifically mixtures of 50% PG94-10 with 50% PG76-22 and 60% PG94-10 with 40% PG76-22. In addition to these blends, the research explored two laboratory-modified binders to address the limitations of conventional PG and blended binders. These highly modified binders were based on PG64-22 (APAC Southern Missouri, Springfield, MO, USA) and PG76-22, each enhanced with Styrene–Butadiene–Styrene (SBS) polymer modifiers (Kraton Inc., Houston, TX, USA)—8% for PG64-22 and 6% for PG76-22—and 0.1% Sulfur as a cross-linker [14].

Several rheological tests on modified and blended PG binders were performed in conformity with the AASHTO M320-23 [15]. The tests included the Bending Beam Rheometer tests, conducted in conformance with AASHTO T 313 [16] to estimate the low-temperature stiffness, S(t), and m-values of the binders. DSR tests were conducted according to AASHTO T 315 [17] to determine dynamic shear modulus, Complex modulus (G*), and phase angle (δ). Long-term aging simulation was performed based on AASHTO R 28 [18], and short-term aging simulation was conducted based on the specification given by AASHTO T 240 [19]. The results of these tests, conducted to provide a comprehensive performance grade (PG) overview for each binder, are summarized in

Table 2. Performance grade (PG) blended and modified binders.

Binder	Performance Grade (PG)
50% PG94-10 + 50% PG76-22	PG82-16
60% PG94-10 + 40% PG76-22	PG88-16
PG64-22 + 8% SBS + 0.1% Sulfur	PG82-22 (PM)
PG76-22 + 6% SBS+ 0.1% Sulfur	PG88-16 (PM)

Note: PM = Polymer Modified.

[20].

2.3. Preparing High Friction Surface Treatment Samples (Coupons and Slabs)

2.3.1. Preparing Aggregate Coupons

The main procedures for making the curved specimens (coupons) used in this investigation are shown in Figure 3. To create a smooth, even surface, ready-mix plaster was first applied to the bottom of the molds. To maximize the performance of the HFST samples, aggregates were carefully incorporated into the plaster to ensure uniform distribution and compaction. Binders were poured over the embedded aggregates at a regulated rate to guarantee uniform coverage and a strong bond [21]. After that, the coupons were left to cure, which strengthened the binders and produced a long-lasting bond with the aggregates. The plaster was removed to reveal the prepared surface for testing after it had cured.

2.3.2. Preparing HFST Slabs

To guarantee uniformity and consistency, plywood substrate was utilized during the sample preparation process. Depending on whether modified or PG binders were being used, the binder had to be heated to the proper temperature at first. To guarantee total surface coverage, the heated binder was then evenly applied to the prepared slab at a rate of 0.30 to 0.38 gallons per square yard. The coverage rate required to achieve at least 50% embedment of aggregates is highly dependent on the aggregate gradation and the substrate characteristics. As shown in Figure 3, the heated aggregate was broadcast over the binder at a rate of 12–15 lb/SY, continuing until the point of rejection or refusal was reached or until the binder was fully covered [9,21]. The aggregate was uniformly embedded using a manual compactor and a heavy plate, guaranteeing that the samples operated consistently and optimally. The surface of every slab was rubbed with a wooden board to loosen the material after the sample had dried.

2.4. Aggregate Durability and Performance Testing

The aggregates were tested for durability and performance using the Micro-Deval (MD) apparatus to determine their resistance to degradation and polishing. This test was designed to assess aggregate durability and resistance to polishing and abrasion in the presence of water. Previous research has shown that the MD test has good multi-laboratory accuracy and is a suitable replacement for the soundness tests [22].

The MD test was conducted in accordance with ASTM D6928-17 [23] for coarse aggregates with sizes ranging from 3/8 to #4. The test durations were set at 105, 180, and 240 min, with each aggregate tested at the specified intervals. All aggregate samples were tested using the Aggregate Image Measurement System (AIMS2) manufactured by American Pine Instrument Company (Grove City, PA, USA),both before and after MD abrasion (BMD and AMD). AIMS analyzed two sizes (3/8″–1/4″ and 1/4″–#4) for each aggregate to observe changes in texture (TX) and gradient angularity (GA) indices after MD polishing. GA indices were calculated using black-and-white images to account for surface irregularities, whereas TX indices were calculated using the wavelet analysis method on grayscale images. The fine aggregate MD test followed ASTM D7428-15 [24] on HFST size gradation for all aggregates. The test was conducted for 15 and 30 min, with two samples per run time, and included calculations of mass loss percentage. All weights in this test were based on oven-dried aggregates.

2.5. Performance Tests for Friction Properties

To evaluate the frictional characteristics and the effects of various aggregate types and sizes in asphalt-based HFST applications, a wide range of tests was used. The suitability of binder and aggregate sizes for HFST was determined through these performance tests. The British Pendulum Test (BPT), which was conducted with the Accelerated Polishing Machine (British Wheel), was used for preliminary evaluation of the aggregate–binder interactions and to assess the skid resistance of coupons before and after 10 h polishing. The Dynamic Friction Tester (DFT) was utilized in conjunction with the NCAT (Auburn, AL, USA) Three-Wheel Polishing Device (TWPD) to measure the frictional properties of the pavement surface under traffic-simulation conditions. The Circular Track Meter (CTM) was also used to evaluate the pavement’s macrotexture and texture depth. These tests, shown in Figure 4, provided valuable insights into selecting optimal aggregates in terms of gradation and size for asphalt-based HFST applications, contributing to enhanced pavement performance and safety.

2.5.1. British Pendulum Tester (BPT)

The BPT is widely used in a variety of research domains, including applications based on the AASHTO T 278 standard [25]. In this study, the BPT was used to assess and screen the preliminary influence and performance of aggregate gradation on asphalt-based samples. It measured the frictional forces created by a rubber pad sliding across the sample surface before and after polishing. Initially, at least five tests were performed on each sample to establish baseline British Pendulum Number (BPN) values. Following these pre-polishing tests, the aggregates in the samples were polished with a British wheel. The BPN values were then measured again with the BPT to determine the post-polish BPN values.

2.5.2. Accelerated Polishing of Aggregates Using the British Wheel

In compliance with AASHTO T 279-18 standard [26], the aggregate coupons were polished using the British wheel following the initial testing with the British Pendulum Tester (BPT). Fourteen aggregate coupons were firmly fastened around the wheel’s perimeter during each test run. The wheel was turned at 320 ± 5 rpm all the time. Concurrently, a pneumatic tire wheel was positioned to exert pressure on the aggregate coupons’ surface. The total load applied during this process was 391.44 ± 4.45 N. This configuration made it possible to perform a controlled polishing process that mimicked the wear patterns seen on real road surfaces.

2.5.3. Dynamic Friction Test (DFT) and Three-Wheel Polishing Device (TWPD)

The DFT as described in ASTM E 1911-19 [27], consists of a circular disk with three rubber slider pads that are able to rotate at rates of up to 100 km per hour. Once the disk reaches the target speed, it is put onto the pavement surface, where the Coefficient of Friction (COF) is measured as it gradually decelerates. Friction measurements are taken in wet conditions, with the findings averaged from two repetitions to assure accuracy. The TWPD mimics traffic wear on the samples based on the guidelines outlined in AASHTO PP 104-21 [28]. This machine includes a turntable with three pneumatic rubber wheels and a water spray system to simulate wet conditions. The technique reduces rubber tire wear and removes surface particles, allowing for additional polishing. The COF is measured at various phases of the polishing process: at the start (0 cycles), then at 30,000, 70,000, and 140,000 cycles, providing a complete assessment of friction performance throughout the polishing stages.

2.5.4. Circular Track Meter (CTM)

The CTM measures the Mean Profile Depth (MPD) at a fixed location, following the guidelines set out in ASTM E2157 (2019) [29]. The MPD quantifies the average depth of the pavement surface texture; it is fitted with a charge-coupled device (CCD) laser-displacement sensor positioned on a spinning arm. This design allows the sensor to move over a circular track with a diameter of 284 mm (11.2 in). This enables comprehensive measurement of pavement texture and macrotexture, providing valuable data for evaluating surface characteristics under various conditions [30,31,32,33,34,35].

3. Results and Discussions

To examine the performance of several HFST samples that included different binders and aggregates of varying sizes, the experimental test results were thoroughly investigated. Various tests were conducted as part of the investigation to look into how aggregate properties affected the effectiveness of asphalt-based HFSTs. These tests included performance evaluations utilizing a variety of techniques as well as durability assessments of aggregates with varying gradations. In particular, skid resistance was measured using the British Pendulum Test (BPT), and wear was simulated by accelerated polishing using the British wheel. The DFT offered more information on COF, and the CTM was used to quantify MPD.

3.1. Micro-Deval (MD) and Aggregate Image Measurement System (AIMS) Results

Figure 5 displays the results of the MD test, which was performed in accordance with ASTM D6928 for coarse and medium aggregate sizes and ASTM D7428 for HFST-sized aggregates. The data show that as MD polishing time increased, all gradations experienced higher percentages of mass loss. The HFST-sized aggregates had a lower mass loss, most likely due to differences in polishing times and test methods. CB had the greatest mass loss of the coarse and medium gradations, while Meramec had the least. Rhy and Flint had similar mass loss percentages for polishing cycles AMD 105, 120, and 180. However, at AMD 240, Rhy showed a lower mass loss than Flint. For the HFST size, Meramec had the lowest mass loss after 30 min of abrasion time, followed by CB, with Rhy and Flint showing similar results.

The AIMS device proved valuable for quantifying the shape, angularity, and surface texture of aggregates. In this study, it assessed particle shape for four aggregate types in two sizes before and after abrasion in the MDA (105, 180, and 240 min). Through the utilization of the AIMS device, the surface texture properties of these friction aggregates were accurately evaluated before and after polishing. Figure 6 and Figure 7 illustrate the AIMS texture and angularity indices for different aggregates of two sizes: 3/8″–1/4″ and 1/4″–#4. These indices were measured before MD polishing (BMD) and after MD polishing at various times (AMD 105, AMD 180, and AMD 240). The reduction in aggregate size from 3/8″–1/4″ to 1/4″–#4 resulted in a decline in texture indices across all phases: BMD, AMD 105, AMD 180, and AMD 240. Notably, CB exhibited higher texture indices for AMD 240 in the 1/4″–#4 size range compared to the 3/8″–1/4″ size range.

Additionally, CB, Rhy, and Flint showed decreasing texture trends with increasing polishing time for aggregates of size 3/8″–1/4″. However, for the 1/4″–#4 size, their texture initially decreased and then increased at AMD 240. In contrast, Meramec exhibited an increase in texture for both sizes up to AMD 105, followed by a decrease in texture.

Figure 7 illustrates the angularity index for various aggregates across different polishing times, revealing different trends. Specifically, for CB and Rhy, the angularity decreased with increasing polishing time. In contrast, Flint exhibited a more variable pattern: angularity decreased after 105 min, increased after 180 min, and then decreased again after 240 min of polishing. Meramec showed a different trend, with angularity initially decreasing after 105 min, increasing slightly after 180 min, and then rising further after 240 min of polishing. The variations in texture and angularity at different polishing times may be attributed to the MD polishing process, during which some aggregate particles may have broken, exposing their internal surfaces. This increased surface roughness can contribute to higher texture and angularity indices. Additionally, the polishing process might have uncovered previously hidden textured surfaces, leading to an increase in both texture and angularity as these rougher surfaces become more prominent.

3.2. Effect of Different Aggregate Size and Binders on the British Pendulum Number

The BPN values obtained between pre- and post-ten-hour polishing cycles using the British Wheel with three distinct aggregate sizes (HFST, medium, and coarse) and binders are shown in Figure 8. Two coupons were made for each aggregate size using different PG and modified binders, as shown in Table 2, with Epoxy resin serving as the control binder. For every aggregate coupon, five BPN measurements were made both before and after polishing; the average BPN values were then computed.

CB had the highest BPN values for the original HFST size among all aggregates, both before and after polishing. Rhy and Flint also produced comparable results. At first, BPN values were similar in every sample. Nevertheless, samples containing PG and modified binders showed a greater decrease in BPN levels after polishing for 10 h. There were some exceptions, including Rhy and Meramec. For example, after polishing with PG88-16, the BPN value for Rhy in the HFST size decreased by only 12%, which was less than the 17% decrease observed with Epoxy. For medium-sized aggregates, when switching from Epoxy to asphalt-based binders, a more noticeable decline in post-polish BPN values was observed compared to the HFST size.

In general medium-sized aggregates in CB, Rhy, and Meramec showed a less pronounced decline in BPN after polishing compared to the HFST-sized aggregates. For instance, with CB and PG82-16, the BPN value for the HFST size decreased by 14%, whereas the medium size saw a smaller reduction of around 10%. In contrast to the HFST size, Flint showed a higher decrease in BPN values, particularly when using PG binders. In comparison to HFST and medium-sized aggregates, the observed trend for coarse-sized aggregates shows a more moderate decrease in BPN values across all binders. Following the polishing process, Epoxy once more showed the least amount of BPN decrease, followed by PG binders and modified binders. The results indicate that aggregate size may be important when using PG and modified binders, as CB and Flint showed a smaller BPN decrease with PG88-16 compared to Epoxy, and Rhyolite exhibited a lower BPN reduction after polishing with PG82-16, with decreases of 15.5% and 16.5% for PG82-16 and Epoxy, respectively.

3.3. Effect of Different Aggregate Sizes and Gradation with Different Binders on the Coefficient of Friction (COF)

The DFT was used to measure the COF values during the first 0 cycles, 30K cycles, 70K cycles, and final 140K cycles of polishing, all of which were carried out at a speed of 20 km/h. Based on their durability and AIMS results, CB and Rhy were chosen for evaluation among the aggregates; three different sizes were taken into consideration: HFST, medium, and coarse. For the accelerated friction testing, the study also included four binders: Epoxy Resin, PG88-16, PG82-22 (PM), and PG88-16 (PM), which were selected based on how well they performed in BPT tests. For every combination of polishing cycle and speed, a single friction measurement was made using the DFT, and the COF values represent the average of two replicates.

As illustrated in Figure 9, the initial COF values at 20 km/h, before polishing, were comparable across all binders for the HFST gradation, with CB showing higher COF values than Rhy. After 30K polishing cycles, the COF decrease for CB was smaller for Epoxy compared to the other binders. However, as the number of polishing cycles increased, PG82-22 (PM) showed the least decline in COF among the asphalt-based binders. For Rhyolite, after 30K cycles, PG88-16 delivered results comparable to, and occasionally better than, Epoxy resin. Nonetheless, as polishing cycles reached 70K and 140K, the COF values decreased, with Epoxy resin showing the best performance at 140K cycles. For the medium gradation with CB, increasing the number of polishing cycles resulted in a consistent decline in COF, with Epoxy resin outperforming the other binders by exhibiting a lower decrease in COF. Among the other binders, PG88-16 (PM) performed better after 30K polishing cycles, but after 70K cycles, PG82-22 (PM) achieved the highest COF values. In the case of Rhy, PG82-22 (PM) exhibited higher COF values over 30K cycles. However, as the polishing cycles increased to 140K, Epoxy resin again demonstrated the highest COF, closely followed by PG88-16 (PM).

In the coarse gradation, CB once again exhibited higher COF values than Rhy before and after polishing. However, following polishing, the COF values for coarse aggregates differed from those of the original HFST and medium-sized aggregates. Slabs made with asphalt-based binders showed COF values comparable to or even higher than those of the Epoxy-based slabs. For CB, the PG88-16 binder exhibited the smallest reduction in COF after polishing for 140K cycles, with a decrease of only 19%. In contrast, Epoxy showed a reduction of 23%, indicating that PG88-16 outperformed Epoxy in maintaining friction after extensive polishing. In the case of Rhy, the initial COF values were about the same and comparable for all binders. After 30K polishing cycles, the COF values remained comparable, with PG88-16 showing slightly higher performance. As the polishing cycles reached 140K, PG88-22 (PM) recorded the highest COF of all the binders tested. For Rhy with PG88-22 (PM), the COF decreased by approximately 38%, whereas the reduction for Epoxy was significantly higher at 77%. This trend shows that larger aggregate sizes contribute to improved friction performance in samples with PG and modified binders as alternatives to Epoxy resin.

3.4. Effect of Aggregate Size and Gradation with Different Binders on the MPD (mm)

The Circular Track Meter (CTM) was used to measure the slabs’ surface macrotexture in order to determine the MPD. Figure 10 shows the MPD measurements both before and after multiple polishing cycles for different combinations of slabs made with CB and Rhy, each with three distinct gradations (HFST, medium, and coarse). Compared to the other binders, Epoxy resin’s change in MPD slope by increasing the polishing cycles was less noticeable, resulting in more stable texture retention. Interestingly, with the increase in aggregate size, the reduction in MPD became more stable and linear, especially for the PG82-22 (PM) binder. For example, for CB in HFST size, once polishing cycles went up to 140K, about a 29% reduction in MPD was observed, while for the coarse CB, this went down to about an 8% reduction. Similarly, in the case of Rhy for HFST size, the MPD decreased by about 41%, whereas with the increase in gradation to coarse, the percent decrease was reduced to 31%. This trend suggests that larger aggregates contribute to improved performance during extended polishing cycles.

4. Conclusions

High Friction Surface Treatments (HFST) are the first pavement treatment recommended for safety because of their potential to improve skid resistance and prevent traffic accidents. While Epoxy-based HFSTs are widely used, they face several challenges, including compatibility with existing pavements, high installation and removal costs, and durability concerns for existing pavements with marginal substrate conditions. This study investigated the effects of aggregate size, type, and gradation on the performance of asphalt-based HFSTs.

Several durability and performance friction tests were performed; among them, the MD and AIMS tests for determining the characteristics and durability of the aggregates and the BPT, British Wheel Test, and DFT for determining values of COF before and after polishing with TWPD. In this respect, the MPD and the macrotexture were determined by a CT-Meter.

The result showed that the choice of aggregate size and gradation greatly influences asphalt-based HFST performance as an alternative for Epoxy. In the MD test, Meramec showed the least mass loss amongst all sizes and aggregates, meaning that it had the highest durability and abrasion. In coarse-sized aggregates, Rhy showed the second least after Meramec, having only a reduction of 6% after 240 min of abrasion. The CB ranked as the second lowest among the HFST size categories in mass loss of about 4% after Meramec. From the AIMS results, there was a reduction in texture indices within the aggregate size reduction from 3/8″–1/4″ to 1/4″–#4. CB displayed higher AMD 240 texture indices within the 1/4″–#4 size range than in the 3/8″–1/4″ size range; hence it is capable of holding texture and frictional properties through polishing and traffic.

The BPT was used to assess the preliminary performance of aggregate gradation on asphalt-based samples. Among all the aggregates, CB had the highest BPN values both before and after polishing, while Rhy and Flint gave similar performances. All samples showed similar BPN values at the beginning. After 10 h of polishing, samples with PG and modified binders showed a higher reduction in the BPN level. The amount of BPN lost by asphalt-based HFST decreased with increasing aggregate size. There were some minor exceptions, but overall, aggregate gradation was critical when asphalt-based binders were used. Aggregates of medium size in CB, Rhy, and Meramec demonstrated less pronounced loss in BPN after polishing compared to those of HFST size. For example, with CB and PG82-16 binders, the HFST size BPN value decreased by about 14%, while in the case of a medium-sized aggregate, this decrease was about 10%. Compared to coarse-sized aggregates, the trend observed for HFST and medium-sized aggregates indicated a more gradual decrease in BPN values. As Rhyolite had a reduced BPN after treatment with PG82-16, the respective percentage decreases were 15.5% and 16.5% for PG82-16 and Epoxy.

The results of the DFT of COF for the various slabs tested showed that at 20 km/h before polishing, the COF values were roughly identical for all binders, with CB having a higher friction value than Rhy. The COF in asphalt-based slabs decreased more than in Epoxy-based slabs as polishing cycles increased for HFST and medium gradations. However, in the coarse gradation, after polishing cycles, the COF for slabs made with modified binders and PG binders was equal to or higher than that for Epoxy. For CB, the PG88-16 binder showed the smallest reduction in COF after 140K polishing cycles, reducing by only 19%, while Epoxy showed a reduction of 23%, indicating that PG88-16 outperformed Epoxy in maintaining friction after extensive polishing. In the case of Rhy, as polishing cycles reached 140K, PG88-22 recorded the highest COF among all binders tested, with a decrease of about 38%, while the reduction for Epoxy was significantly higher, at 77%.

The MPD results for the slabs indicated that Epoxy resin consistently provided the highest MPD values for both aggregates with varying gradations as polishing cycles increased. The change in slope of the MPD for Epoxy resin in both aggregates was less pronounced when compared to other binders. Notably, as aggregate size increased, the slope of MPD decreased with more polishing cycles, becoming more stable and linear. This trend became more pronounced in PG82-22 (PM), indicating improved performance with larger aggregates as the polishing cycles progressed.

These findings indicate that due to the differences in stiffness, bonding mechanisms, and other characteristics between asphalt and Epoxy, the durability and performance of asphalt-based HFSTs can be enhanced by selecting the appropriate aggregate type and size. Notably, using larger aggregate sizes significantly improves friction stability, underscoring the importance of utilizing various aggregate gradations to optimize HFST applications and using them according to the section with the need to improve friction with marginal conditions.