Abstract
This study investigates the fatigue and healing performance of asphalt mixtures incorporating nanoparticle-modified binders, using nanoclay and nanosilica. Two binders (PG 58-28 and PG 76-22) were selected to assess their effect on healing behavior. Asphalt mixtures were prepared using granite aggregates and UFGS Gradation 3, then subjected to mechanical and simulation-based evaluations. Cyclic fatigue testing was conducted using the asphalt mixture performance tester at 25°C, with a strain amplitude of 800 microstrains and a loading frequency of 10 Hz. To simulate in-service conditions, 10- and 20-minute rest periods were introduced after 25% of the specimen’s estimated fatigue life, representing early stage fatigue damage accumulation. Healing was quantified by comparing the number of cycles to failure (Nf) before and after rest period. Dynamic modulus testing and FlexPAVE™ simulations were also performed to assess viscoelastic behavior and long-term pavement performance. Results showed that both nanoclay and nanosilica modified mixtures exhibited notable improvements in fatigue life relative to the control, with nanoclay modified mixtures achieving the highest fatigue life improvement, while nanosilica modified mixtures demonstrated consistent intermediate gains across both binder types. FlexPAVE™ simulations indicated a 37% reduction in total fatigue damage over 20 years for nanoclay-modified mixtures with rest periods. Rutting and cracking resistance also improved significantly, as observed from indirect tensile asphalt cracking test and asphalt pavement analyzer tests. The findings confirm that nanomodification, especially with nanoclay, enhances the intrinsic healing capacity, fatigue resistance, and durability of asphalt mixtures. Incorporating rest periods in design further optimizes long-term performance, offering a sustainable strategy for modern pavement systems.
Introduction
Fatigue cracking is one of the primary distresses in asphalt pavements ( 1 ). Such cracking is caused by horizontal tensile strains occurring at the bottom of asphalt layers, generated because of repeated traffic loading on aged pavements ( 2 – 4 ). The reduced flexural capacity of the aged asphalt layer causes crack initiation followed by crack propagation to the top surface ( 5 ). This type of distress shortens the pavement’s service life, requiring more frequent repairs and contributing to elevated maintenance expenses and higher greenhouse gas emissions because of repeated asphalt production and application ( 6 – 9 ). Between 2009 and 2017, expenditures related to crack-related pavement maintenance rose significantly, reaching as high as $21.4 billion, highlighting the substantial economic burden associated with cracking ( 10 , 11 ). Studies indicate that the incorporation of healing additives to asphalt can repair microcracks resulting in extended service life ( 6 , 12 ). Healing in asphalt binders refers to their capacity to regain lost mechanical qualities, such as stiffness or tensile strength, when subjected to rest periods ( 13 , 14 ). Nonetheless, conventional asphalts’ natural capacity for healing is insufficient to completely seal microcracks or restore structural integrity, especially when subjected to repetitive loadings and therefore requires external stimuli (extrinsic methods) ( 15 ) or binder modification ( 16 ) to significantly improve the healing performance and ensure long-term pavement durability.
Extrinsic healing methods such as induction heating, microwave heating, microencapsulation, and hollow fiber systems are widely studied to enhance asphalt pavements’ durability ( 17 , 18 ). Induction heating involves the addition of steel fibers to asphalt mix to enable localized heating through alternating magnetic fields which soften the binder, allowing cracks to close. However, this method may result in accelerated binder aging and reduced mechanical performance ( 19 ). Microwave heating emerged as a simpler alternative, eliminating the need for conductive additives and reducing material costs but the concerns remain over health risks, as microwave radiation can cause tissue damage if safety measures are not properly implemented ( 20 ). Researchers have explored the use of rejuvenators to promote crack healing in asphalt mixtures, primarily utilizing two delivery methods: microencapsulation and hollow fiber embedding ( 21 , 22 ). The stress concentration at the crack tips causes the microcapsules to rupture, releasing the rejuvenator into the crack, where it gradually penetrates and fills the damaged area, promoting the repair of microcracks and facilitating crack healing ( 23 ). However, the drawbacks such as uncontrolled softening ( 24 ) and agglomeration of microcapsules have restricted the use of this technique ( 25 ).
Molecular-level modification of asphalt is considered the most effective approach for enhancing the self-healing capacity of asphalt mixtures ( 26 ). Strengthening the material’s intrinsic healing properties offers a promising alternative to traditional maintenance methods by allowing internal damage recovery without extensive surface intervention. This approach not only reduces long-term maintenance costs but also offers environmental and operational advantages, including lower carbon emissions and improved roadway safety compared with traditional repair methods ( 27 , 28 ). Among emerging techniques, the incorporation of nanomodifiers presents a promising approach for developing self-healing asphalt materials ( 29 ). Some of these nanomodifiers include nanoclay, nanosilica, nanoalumina, nano rubber, carbon nanotubes, graphene nanotubes, and nano titanium dioxide amongst others ( 30 , 31 ). Nanoparticles are drawn toward crack tips because of their high surface energy, where they gather to prevent crack propagation and assist in healing the damaged asphalt material and, moreover, the nanoparticles can also improve ageing, and rheological and thermal properties of asphalt mixtures ( 26 ). As a results of its spherical shape, high specific surface area, small particle size, and greater density relative to bitumen, nanoparticles have proven to accelerate molecular random motion, enhance the flow of bitumen binder into microcracks, and ultimately improve the self-healing performance of hot mix asphalt ( 17 ).
The healing of asphalt has been extensively studied by previous researchers using different nanomodifiers and testing protocols. Ganjei and Aflaki ( 32 ) evaluated the self-healing behavior of asphalt mixtures modified with nanosilica and styrene–butadiene–styrene (SBS), concluding that the combined addition of nanosilica and SBS significantly enhances the self-healing capabilities of hot mix asphalt. Badroodi et al. ( 33 ) explored the use of nanosilica to improve the self-healing capabilities of warm mix asphalt (WMA) containing reclaimed asphalt pavement (RAP). The results demonstrated a marked enhancement in the healing performance of the WMA-RAP mixtures with nanosilica incorporation. Amin et al. ( 34 ) improved the self-healing performance of asphalt mixtures by incorporating spherical nanosilica characterized by small particle size and a high density of 2.33 g/cm3. At a 3% nanosilica dosage, the self-healing index achieved a peak value of 88.7%. Tabatabaee and Shafiee ( 35 ) demonstrated that the addition of organoclay into asphalt binders enhanced both the fatigue life and their ability to heal. Xiang et al. ( 36 ) evaluated the influence of healing temperature and duration on the fatigue–healing behavior of both neat and SBS-modified asphalt binders and mixtures using the fatigue–healing–fatigue and 4-point bending beam fatigue test, respectively. They found that SBS-modified systems achieved higher fatigue life recovery than neat binders/mixtures with highest performance at an optimal temperature (50°C for neat, 60°C for SBS) and optimal healing time (30 min for binders, ≈12 h for mixtures). Oliveira et al. ( 37 ) evaluated the impact of rest periods (4-hour intervals) on stiffness and fatigue life in asphalt mixtures with 5% fly ash-modified samples using dynamic complex modulus and cyclic fatigue tests. Key findings showed up to 90% stiffness recovery during rest, but no significant extension in fatigue life (Nf) with no significant influence of fly ash addition on the performance.
Among the various test methods used to evaluate healing, recent studies have compared their effectiveness and limitations. Zeiada et al. ( 38 ) compared beam and cyclic fatigue tests and found the uniaxial test to be more effective for assessing healing behavior. While the beam test yielded longer fatigue lives, it exhibited higher variability because of specimen geometry and complex flexural stresses and showed limited sensitivity to healing. In contrast, the cyclic test allowed faster testing, more repeatable specimen preparation, and clearer assessment of damage progression. Notably, healing measured in the uniaxial test was over 10 times greater than in the beam test under similar strain conditions, reinforcing its suitability for capturing intrinsic healing in nanomodified mixtures.
While nanoparticles have exhibited notable potential in enhancing the intrinsic healing ability of asphalt binders, the underlying mechanisms remain largely unexplored, and limited data exist concerning their long-term effects on asphalt mixture performance. To bridge these knowledge gaps, this study investigates the fatigue recovery and healing behavior of asphalt mixtures modified with nanoparticles, using controlled damage levels and intermittent rest periods.
Building on the authors’ prior binder-level investigation by Chaudhary et al. ( 16 ), which systematically examined nanomodifiers dosages of 1%, 3%, and 4%, under varying rest periods (10, 20, and 30 min) and damage levels (25%, 37.5%, and 50%) using Linear Amplitude Sweep (LAS) and Linear Amplitude Sweep Healing (LASH) protocols. The findings demonstrated that the 4% dosage consistently yielded the highest healing index and fatigue life recovery particularly at the 20-minute rest period and 25% damage level while lower dosages showed smaller or inconsistent improvements. Based on these experimentally validated results, the same dosage was extended to the mixture-level evaluation in the present study to maintain consistency and to isolate the influence of material type and rest period on healing performance. This phase of the study specifically assesses the extent to which the improvements observed at the binder level translate into enhanced healing and fatigue resistance at the mixture level under simulated traffic loading using a modified cyclic fatigue test. By introducing rest periods between loading cycles in controlled laboratory fatigue tests, the study captures the intrinsic healing behavior of nanomodified asphalt mixtures under conditions that better reflect in-service pavement operations. To further validate and extend the laboratory findings, FlexPAVE™ simulations were conducted to model the long-term damage accumulation and healing potential of the mixtures under field-representative conditions, including realistic traffic loads, pavement structure, and environmental factors. This underscores the practical relevance of incorporating rest periods and nanomodification into modern pavement design frameworks to enhance long-term performance, reduce maintenance needs, and promote sustainable infrastructure development.
Objectives
The primary objective of this study is to examine the healing behavior of asphalt mixtures incorporating nanoparticle-enhanced binders. The specific objectives of the study can be outlined as follows:
Assess the fatigue and healing capabilities of asphalt mixtures containing nanoparticle-modified binders by analyzing the effects of rest periods on fatigue recovery under controlled loading scenarios
Explore how different nanomodifiers (nanosilica and nanoclay), each at a 4% dosage, influence the mechanical performance of asphalt mixtures, including improvements in fatigue life, rutting resistance, and crack mitigation.
Evaluate the long-term performance of nanomodified asphalt mixtures using FlexPAVE™ and assess their impact on pavement design by quantifying cumulative damage reduction and healing effectiveness over a 20-year service life to simulate real field conditions
Description of Materials
This study uses two performance-graded asphalt binders (PG 58-28 and PG 76-22), two nanomodifiers (nanosilica and nanoclay) for the preparation of nanomodified binders and mixtures. PG 58-28, a conventional (neat) binder, is commonly used in cold-climate pavement applications, while PG 76-22 is a polymer-modified binder designed for enhanced performance in high-temperature environments. These binders were selected to evaluate how binder type influences the healing behavior of nanomodified systems across a broad range of rheological characteristics. Table 1 summarizes the properties of the binders used in the study.
Rheological Properties of Asphalt Binders
Note: BBR = Bending Beam Rheometer; PAV = Pressure Aging Vessel; PG = performance grade; RTFO = Rolling Thin Film Oven; NA = not available.
Nanomodifiers
Two types of nanomodifiers namely nanosilica (SiO2) and nanoclay/nanomontmorillonite ((Na,Ca)0.33(Al,Mg)2(Si4O10)(OH)2·nH2O) were incorporated into the asphalt binders. The nanomodifiers were dispersed using a three-stage shear blending process (200 rpm–5,000 rpm–200 rpm for 30 min each at 160 ± 5°C) ( 10 ). Nanosilica, a highly stable metal oxide with fine particle size and large specific surface area, was selected for its ability to enhance binder stiffness, improve filler dispersion, and increase resistance to oxidative aging, thereby contributing to a more robust mastic structure and improved fatigue resistance at the mixture level ( 26 ). The nanosilica has the molecular formula SiO2, with a grain size of approximately 80 nm, a melting point of 1710 °C, and a molecular weight of 60.09 g/mol. It possessed a specific gravity of 2.4 g/cm3 and contained 0.1% adulterants on a metal basis. The pH of the nanosilica was 6, and it appeared as a white powder.
Nanoclay, known for its layered silicate structure comprising an octahedral alumina sheet sandwiched between two tetrahedral silica sheets and organophilic nature, promotes strong binder-filler interactions and has shown potential to enhance crack-bridging and healing capabilities ( 35 , 39 ). The nanoclay consist of 35–45 wt.% montmorillonite modified with dimethyl dialkyl (C14–C18) amine, with a particle size of ≤ 20 μm, bulk density ranging from 200 to 500 kg/m3, and a moisture loss after drying of ≤ 3%. It appeared as a beige powder and had a basal spacing of 2.39 nm.
Preparation of Nanomodified Asphalt Binders
The binder modification process involved gradual incorporation of nanomodifiers into base asphalt binder preheated to 160 °C. The blending process adopted consisted of initially using a low-shear mixing at 200 rpm for 30 min, followed by 5000 rpm using a high-shear mixer for 30 min and the final low shear stage at 200 rpm for another 30 min. This approach was adopted based on the methodology reported by Zhang et al. ( 40 ) and followed by Chaudhary et al. ( 41 ). The staged mixing sequence promotes effective nanoparticle wetting and dispersion, facilitates agglomerate breakdown during high shear, and stabilizes the modified binder during the final low-shear stage contributing to improved storage stability by limiting nanoparticle agglomeration and reducing the potential for phase separation as reported in the author previous study by Mitra et al. ( 10 ).
Preparation of Asphalt Mixtures
The prepared nanomodified binders were added to the aggregates for the preparation of the mixtures. Asphalt mixtures prepared with nanomodified binders were designed using Unified Facilities Guide Specifications (UFGS) Gradation 3 as shown in Figure 1 (9.5 mm nominal maximum aggregate size, 75 gyrations). The optimum binder content of 5.6% was determined based on achieving target volumetric properties in accordance with Superpave mix design standards, including air voids (4 ± 1%), minimum voids in mineral aggregate of 15%. Following mix design, 150 mm × 180 mm Superpave gyratory compactor specimens of control and nanomodified mixtures were compacted to achieve 7.0% air voids. From the central, uniformly compacted zone of each specimen, four cylindrical cores (38 mm × 110 mm) were extracted using a precision coring drill. Specimen trimming and dimensional verification were performed in accordance with AASHTO PP 99 to ensure compliance with tolerance requirements for flatness and perpendicularity. The viscoelastic and fatigue properties of cored asphalt mixtures were evaluated using both dynamic modulus (|E*|) testing and cyclic fatigue testing. The control and nanomodified asphalt mixtures were also tested to ascertain the cracking resistance and rutting behavior using the indirect tensile asphalt cracking test (IDEAL-CT) and asphalt pavement analyzer (APA) test. All the samples were tested in triplicates to ensure sufficient repeatability

Gradation curve for 9.5NMAS for UFGS Type 3.
Experimental Program
The asphalt mixtures used in this study were prepared with nanoclay- and nanosilica-modified binders at a dosage of 4% by binder weight, following a high-shear mixing protocol. Aggregates conformed to UFGS Gradation 3 with a 9.5 mm NMAS and were compacted to 75 gyrations. Full details of the materials, binder modification, and mixture design are provided in the Materials section.
Figure 2 illustrates the overall experimental framework developed to evaluate the mechanical performance and healing behavior of the nanomodified mixtures. Following compaction, specimens were cored and cut to meet the dimensional requirements of the various laboratory tests. Cyclic fatigue testing was conducted in accordance with AASHTO TP 133 in the asphalt mixture performance tester (AMPT) to assess the influence of nanomodification and rest periods on fatigue and healing performance. Testing was carried out at 25°C using a loading strain of 800 microstrain and a frequency of 10 Hz. After inducing 25% of the number of cycles to fatigue (Nf), rest periods of 10 and 20 min were introduced. The selection of 25% damage level was based on findings by Kim et al. ( 42 ) and Baglieri et al. ( 43 ), which reported that nanomodifiers are most effective when damage exist as microcracks, making it ideal for healing. This threshold aligns with Chaudhary et al. ( 16 ) binder-level study, which used 10- and 20-minute rest periods and 25% damage. Similar rest durations were used by Mannan et al. ( 44 ) and Xiang et al. ( 36 ) These parameters reflect realistic pavement behavior, where early stage fatigue and intermittent traffic loading allow rest-induced healing to occur. Healing potential was quantified by comparing the number of cycles to failure (Nf) before and after rest period. Dynamic modulus testing was carried out using the AMPT to characterize the viscoelastic response of the mixtures across multiple temperatures and loading frequencies. These data were also used as input for FlexPAVE™ simulations. The cracking resistance of the mixtures was assessed using the IDEAL-CT, while rutting susceptibility was evaluated using APA. These tests provided additional insight into how nanomodifiers influence resistance to common pavement distresses. Lastly, FlexPAVE™ simulations were used to evaluate long-term performance under field-representative traffic and climate conditions, enabling the quantification of cumulative damage and healing over a 20-year service period. Further details of the test methods and parameters are provided in subsequent sections.

Experimental methodology.
Cyclic Fatigue Test (AASHTO TP 133)
The cyclic fatigue test was performed to evaluate fatigue resistance and damage evolution in asphalt mixtures. For each mixture type, three replicate specimens of dimensions 38mm × 110 mm were fabricated and tested, maintaining a target air voids content of 7±0.5%. To ensure that fatigue life fell within the acceptable range of 2,000 to 80,000 cycles, as specified in AASHTO TP 133, preliminary trials were conducted using varying strain levels. Based on these trials, a strain amplitude of 800 microstrain was selected, as it consistently produced failure cycles within the standard-specified window and ensured valid test conditions. The test was then performed using the AMPT in strain-controlled mode, applying sinusoidal loading at a frequency of 10 Hz at 25°C. The cycle where the measured phase angle drops sharply or the product of the peak to peak and cycle number reaches a maximum value was considered as the failure cycles (Nf) as per AASHTO TP 133. To assess the influence of rest-induced healing, a fatigue rest fatigue testing protocol was implemented, introducing the rest periods (10 and 20 min) interval after the specimen reached 25% of its fatigue life (Nf). This fatigue rest fatigue sequence was designed to capture the intrinsic healing response of the asphalt mixtures. Only specimens that failed within the designated 70 mm central gauge length monitored using linear variable differential transformers were considered valid for analysis. Failures occurring outside this region were excluded to eliminate potential inaccuracies arising from boundary effects, misalignment, or uneven load distribution.
Asphalt Pavement Analyzer (APA, AASHTO T340)
The APA test was performed to evaluate the rutting potential of both control and the nanomodified mixtures in accordance with AASHTO T340. This evaluation was critical to ensure that enhancements in healing properties through nanomodification did not compromise the binder’s ability to resist permanent deformation under repeated loading. Specimens were compacted to meet the APA dimensions of 75mm × 150 mm and a 7% target air voids level. The test was performed at a controlled temperature of 58°C, consistent with the high-temperature performance grade of PG 58-28 binders used in this study and the test samples were preconditioned at the testing temperature for 6 h minimum before testing. The APA machine was configured to apply a hose pressure of 690 ± 35 kPa (100 ± 5 psi) and a wheel load of 445 ± 22 N (100 ± 5 lbf) to each specimen. The experiment was run for 8,000 wheel loading passes in total, or until the maximum rut depth of 14.5 mm was reached, whichever came first. The mean rut depth across all tested specimens is commonly used as the performance indicator.
Indirect Tensile Asphalt Cracking Test (IDEAL-CT, ASTM D8225)
To evaluate the intermediate-temperature cracking resistance of the asphalt mixtures, the IDEAL-CT was conducted in accordance with ASTM D8225-19. For this study, cylindrical specimens were fabricated in the Superpave gyratory to a height of 62 mm and a diameter of 150 mm at a 7% target air voids. The test samples were conditioned at 25°C for a minimum of 2 h and were subjected to a loading rate of 50 mm/min until failure. The crack tolerance index (CTindex) serves as a critical complementary metric to assess whether nanomodified mixtures retain sufficient fracture resistance to withstand intermediate-temperature cracking under cyclic traffic loads ( 45 – 47 ). Multiple cracking-related parameters can be extracted from the load–displacement curve obtained during the test, such as fracture energy (Gf), ductility of the mixture(|I75|), crack propagation rate (|m75|), and CTindex. Better cracking performance for asphalt mixtures is indicated by fracture energy, higher values of CTindex, and ITS ( 21 , 22 ).
Dynamic Complex Modulus Test (|E|*, AASHTO TP 132)
The dynamic modulus test was performed to evaluate the linear viscoelastic behavior and stiffness response of asphalt mixtures. All test specimens were prepared with 38 mm diameter and 110 mm height in accordance with AASHTO PP 99. These test standards recommend testing asphalt mixtures at temperatures of 4, 20, and 35°C when prepared with PG 58-XX binders and softer. For mixes prepared using PG 64-XX and stiffer binders, they are to be tested at 4, 20, and 40°C. The test consists of applying a sinusoidal loading at three frequencies of 0.1, 1, and 10 Hz. The resulting dynamic modulus data were employed to develop master curves applying the time-temperature superposition principle at a reference temperature of 20°C using the sigmoidal model. The master curves represent mixture stiffness across varying loading frequencies and are essential for predicting pavement performance under real-world conditions, serving as key inputs for FlexPAVE™ to simulate long-term damage and fatigue life.
Pavement Performance Simulation Using FlexPAVE
The current study employed FlexPAVE, a viscoelastic continuum damage-based simulation tool developed by the FHWA to extend the binder and mixture-level findings to realistic pavement performance. It enables mechanistic–empirical evaluation of asphalt pavement fatigue performance under realistic traffic loading and climatic conditions ( 48 ). For this study, the simulation was limited to fatigue-induced cracking, and rutting-related damage was not considered. Mixture properties from the dynamic modulus and cyclic fatigue tests served as primary damage inputs, which were transformed into the damage characteristic curve used in the FlexPAVETM framework to model the rate and accumulation of microcrack propagation over time. Pavement structure, traffic loading, climatic conditions, and analysis period (20 years) were held constant across all cases to ensure consistent comparison. The analysis focused exclusively on total damage, which represents the combined effect of top-down and bottom-up cracking throughout the pavement structure. This metric was selected to capture a comprehensive assessment of fatigue resistance across the entire depth of the asphalt layer. To simulate real-world conditions, climate data were obtained from LTPPBindv3.1 for Fairbanks, Alaska, and binder selection followed AASHTO M320-10 and AASHTO M332-14 standards. Since the primary goal was to evaluate the impact of nanomodification and healing, a representative pavement structure was assumed: a 4-inch asphalt concrete (AC) layer, a 10-inch granular base, and an 8-inch subbase. AC modulus inputs were derived from laboratory dynamic modulus tests, while the resilient moduli for the granular layers were selected from the internal material library of FlexPAVETM. By using this simulation platform, the study was able to assess how nanomodified mixtures perform under prolonged loading and environmental cycling.
Results and Discussion
Fatigue and Healing Behavior of Nanomodified Asphalt Mixtures
Figure 3, a and b , illustrates that nanomodification significantly enhances the fatigue resistance of asphalt mixtures, with nanoclay and nanosilica yielding up to 197% and 82% improvement in PG 58-28 mixtures, and 172% and 29% in PG 76-22 mixtures, respectively, without any rest periods (0 min). Nanoclay consistently outperformed nanosilica because of its lamellar structure, which improves load transfer and resists crack propagation under repeated loading ( 37 ). In addition, PG 76-22 mixtures exhibited superior fatigue life over PG 58-28, attributed to the polymer-modified binder’s elastomeric and viscoelastic properties, which enhance tensile strain accommodation and delay crack initiation ( 38 ). As shown in Figure 3b, combining nanoclay with PG 76-22 offers the highest fatigue performance, highlighting the synergistic benefits of binder grade and nanomodification strategies. The fatigue resistance of the nanomodified asphalt mixtures were further accessed after rest periods were introduced. Figure 3, a and b , also represents the percentage increase of fatigue life of nanomodified asphalt mixtures relative to the control mixture at different rest periods. The values annotated on the bars highlight the percentage increase in failure cycles because of nanomodifications under 0-, 10-, and 20-minute rest periods. The percentage increment in Figure 3 was calculated relative to the control mixture at the same rest period. As can be seen in Figure 3a, the nanoclay-modified asphalt demonstrated the highest fatigue life, with 117% and 118% improvements over the control mixture after 10- and 20-minute rest periods, respectively, and nanosilica-modified mixtures also exhibited notable gains of 35% and 53% under 10- and 20-minute rest periods when compared with the control mixtures for the same rest periods. A similar trend was seen in mixtures prepared with PG 76-22 binder (Figure 3b), where the nanoclay modified asphalt mixture exhibited 145% and 123% increase in fatigue life for 10- and 20-minute rest period compared with the control mixtures under the same rest periods and nanosilica modified mixtures showed a percentage increase of 57% and 31% in fatigue life when compared with their control counterpart under the same rest periods. This suggests that nanomodified asphalt mixtures are more capable of prolonging pavement fatigue resistance than the control mixtures, regardless of binder type and rest duration.

Failure cycles of nanomodified asphalt mixtures (nanosilica and nanoclay) compared with the control mixture prepared with (a) PG 58-28 binder, and (b) PG 76-22 binder at 0-, 10-, and 20-minute rest period.
The study extended conventional cyclic fatigue testing by incorporating 10- and 20-minute rest periods once mixtures reached 25% of their initial number of cycles to failure (Nf). This simulated real-world traffic unloading, enabling microcrack closure and viscoelastic relaxation. Figure 4, a and b , represents the effect of rest periods on the fatigue life of control and nanomodified asphalt mixtures. The values annotated on the bars highlight the percentage increase in failure cycles for 10- and 20-minute rest periods compared with the 0-minute rest periods. The percentage increment was calculated relative to each mixture’s corresponding 0-minute rest condition.

Impact of rest periods on failure cycles of control and nanomodified asphalt mixtures prepared with (a) PG 58-28 binder, and (b) PG 76-22 binder. Healing potential is assessed by comparing 10- and 20-minute rest periods to the 0-minute condition.
As can be seen from Figure 4a, the control mixtures showed fatigue life increases of 81% and 114% after 10- and 20-minute rest periods relative to ones under continuous loading (without rest periods). The nanosilica-modified mixtures improved by 33% and 79%, while the nanoclay-modified mixtures recorded 32% and 57% increases at 10 and 20 min, respectively, when compared with their counterparts under continuous cyclic loading. A clear linear trend was observed across all mixtures, with fatigue life increasing proportionally to rest duration, confirming the activation of rest-induced healing. The increase in fatigue life with rest periods for the control mixtures is likely because of the intrinsic healing capability of asphalt binder, which allows partial microcrack closure and stress relaxation during unloading phase thereby contributing to the relatively higher percentage recovery observed in control mixtures. While the control mixture exhibited higher percentage recovery because of its lower baseline fatigue life, nanomodified asphalt mixtures maintained substantially higher fatigue resistance after rest periods. The increase in failure cycles of the nanomodified asphalt mixtures is because of their small particle size, spherical geometry, and high density relative to the bitumen matrix facilitating enhanced molecular motion, promoting binder flow into microcracks and accelerating the healing process ( 13 ).
A similar trend was observed in the PG 76-22 mixtures (Figure 4b), where fatigue life consistently increased with rest duration. After a 10-minute rest period, the control mixture improved by 24%, nanosilica-modified mixtures by 51%, and nanoclay-modified mixtures by 11%. Extending the rest to 20 min led to further enhancements, with the control, nanosilica, and nanoclay mixtures showing increases of 66%, 68%, and 36%, respectively. These consistent trends across both binder grades reinforce the effectiveness of rest periods in activating healing mechanisms and the added benefit of nanomodification in enhancing fatigue resistance.
Influence of Nanomodifiers on Permanent Deformation of Asphalt Mixtures
To evaluate the rutting resistance imparted by the nanomodifiers, the APA test was performed. Figure 5 presents the APA rutting test results for asphalt mixtures with PG 58-28 binder, including control, nanosilica-modified, and nanoclay-modified samples. As shown in Figure 5, the nanoclay and nanosilica modified asphalt mixtures had 26.8% and 14.8% reduction in rut depth compared with the control mix. These improvements are consistent with previous studies ( 49 , 50 ) which reported increased stiffness in nanosilica-modified binders, contributing to enhanced resistance against permanent deformation. Nanoclay has been acknowledged to be organophilic and can intercalate within the asphalt matrix, enhancing structural stability and delaying deformation ( 51 , 52 ).

Results of the asphalt pavement analyzer test of the control and nanomodified asphalt mixtures.
Cracking Resistance of Nanomodified Asphalt Mixtures
Figure 6 presents the CTindex values derived from the IDEAL-CT tests for the control and modified asphalt mixtures prepared with PG 58-28 binder. The nanoclay-modified mixtures exhibited the highest CTindex, with an approximate 125% increase with respect to control, while the nanosilica-modified mixtures showed an increase of approximately 74% over the control mixture. The control mixture, in contrast, exhibited the lowest CTindex, underscoring its lower resistance to cracking under loading conditions. This is because nanomaterials enhance asphalt cracking resistance by improving binder cohesion and aggregate adhesion ( 50 ). The results clearly demonstrate that the incorporation of nanomaterials significantly improved the cracking resistance of asphalt mixtures. To understand the toughness and the ductility behavior of the nanomodified asphalt mixtures, the performance interaction charts were plotted (Figure 7). Developed by NCAT, these diagrams plot fracture energy (Gf) against the ductility-to-crack propagation ratio (l75/|m75|), offering insights into a mixture’s toughness (Gf) and flexibility(l75/|m75|). Mixtures with higher Gf and l75/|m75| values, indicating superior toughness and ductility, appear in the upper-right corner of the charts and correspond to higher CTindex values, reflecting better cracking resistance. The dotted contour lines connect points of equal CTindex, facilitating comparisons between mixtures with varying Gf and I75/|m75| combinations but similar overall cracking performance. As shown in Figure 7, nanoclay modified asphalt mixtures exhibited a higher fracture energy and higher ductility (I75/|m75|) ratio as compared with the nanosilica modified and control mixtures. The observation shows that the nanoclay modified asphalt mixtures had a higher toughness and higher ductility than the other mixtures. The nanosilica modified asphalt mixtures also had greater l75/|m75| ratio and higher Gf values than the control; showing that nanosilica improved the toughness and the ductility of the mixture. In addition, as shown in load displacement curve in Figure 8, the nanomodified asphalt mixtures exhibited higher peak load and longer postpeak tail indicating greater fracture energy, higher energy dissipation capacity and higher ductility compared with the control. The nanoclay-modified mixture exhibited the highest peak load and the largest postpeak tail, indicating the greatest fracture energy of 13,800 J/m2 and highest energy dissipation capacity. Nanosilica also improved cracking resistance, achieving a fracture energy of approximately 11,500 J/m2, reflecting moderate gains in ductility and postpeak resistance. In contrast, the control mixture had the smallest curve area, with a fracture energy of 9,200 J/m2, confirming its lower toughness and limited energy-dissipation capability. In summary, the IDEAL-CT test results confirm that nanomodified asphalt mixtures, especially those incorporating nanoclay, exhibit significantly enhanced cracking resistance compared with control mixtures. The improvements in CTindex, coupled with the previously demonstrated benefits in fatigue life and rutting resistance, highlight the multifunctional role of nanomodifiers in enhancing the long-term performance and durability of asphalt pavements. This underscores the potential of nanomaterials as valuable additives in the design of high-performance, crack-resistant asphalt mixtures for sustainable infrastructure applications.

Results of cracking performance of control and nanomodified asphalt mixtures.

Performance interaction chart for CT index.

Load displacement curve for control and nanomodified asphalt mixtures.
Effect of Nanomodification on the Dynamic Modulus of Asphalt Mixtures
Figure 9, a and b , shows the master curves for the control and nanomodified asphalt mixtures prepared with PG 58-28 and PG 76-22 binders, respectively, at a reference temperature of 20°C. It is observed in both figures that nanomodified mixtures showed higher |E*| across the frequency spectrum as compared with the control mixtures, indicating increased stiffness. Specifically, nanoclay-modified mixtures exhibited greater enhancements at low frequencies, suggesting superior resistance to rutting under sustained loading, whereas the impact of nanomodification at high frequencies was less pronounced, showing similar |E*| between nanomodified and control mixtures. These findings indicate that nanomaterials, especially nanoclay can significantly increase asphalt mixtures’ stiffness. This enhancement is primarily because of nanoclay’s organophilic nature, which enables intercalation of the binder between its layered platelets, leading to improved stiffness and structural integrity ( 53 ). The trends were consistent for both PG 58-28 and PG 76-22 binders, confirming that binder grade had minimal influence on the dynamic modulus profiles within this study. These results served as an input for the FlexPAVE simulation to evaluate the effect of nanomodification on the fatigue life of a pavement with and without rest periods.

Master curve for asphalt mixtures prepared with (a) PG 58-28 binder, and (b) PG 76-22 binder.
FlexPAVE Simulation Results: Total Damage after 20 Years
Figure 10 presents the total accumulated fatigue damage for various asphalt mixtures over a simulated 20-year pavement lifespan, as predicted using FlexPAVE™. The damage levels were computed based on mixture properties derived from the dynamic modulus test and cyclic fatigue tests, with and without rest period representing both conventional and modified fatigue loading scenarios. It is observed that the control mixture exhibited the highest damage accumulation confirming its limited ability to withstand long-term fatigue loading. On the other hand, the nanoclay and nanosilica modified mixtures which were subjected to continuous loading (without rest periods) recorded a reduction in the total damage level of 12% and 6%, respectively, relative to the control mixtures. Moreover, the damage mitigation was more pronounced with the incorporation of rest period. The nanoclay modified mixtures after a 20-minute rest period exhibited the lowest damage resulting in a reduction of 13% compared with the control mixtures with rest periods. The nanosilica modified mixture after a 20-minute rest period also improved, showing a 7% reduction in total damage relative to the control mixtures with rest periods. These results reinforce the laboratory findings by validating that nanomodification, especially that nanoclay significantly reduces long-term fatigue damage, and that the inclusion of rest periods further enhances healing behavior. Notably, the nanoclay modified mixtures with the rest period exhibited the lowest total damage, indicating that the combined effect of nanoclay reinforcement and rest-induced healing provides enhanced fatigue resistance under simulated traffic and environmental loading conditions.

Simulated long-term fatigue damage accumulation over a 20-year pavement life for control, nanosilica, and nanoclay-modified asphalt mixtures under conventional (no rest) and healing (20-minute rest) conditions using FlexPAVE.
A comparative analysis of healing benefits from laboratory cyclic fatigue testing and FlexPAVE™ simulations demonstrated consistent trends in rest-induced recovery. The inclusion of a 20-minute rest period resulted in a 114% increase in fatigue life for the control mixture, while nanosilica- and nanoclay-modified mixtures showed increases of 79% and 57%, respectively. These enhancements observed in the laboratory were supported by FlexPAVE™ simulations, which predicted reductions in long-term damage after 20 years of 1% for both the control, 3% nanosilica mixtures, and 2% for the nanoclay mixture. The agreement between experimental and simulated results confirms the reliability of laboratory healing measurements and underscores the potential of incorporating rest periods in pavement design.
Statistical Analysis
The modified cyclic fatigue test showed that fatigue performance of asphalt modified mixtures was influenced by the addition of nanomodifiers. The effect of different variables such as modifier type, rest period, and binder type was observed on the healing of nanomodified asphalt. A multifactor analysis of variance was performed using SPSS software at 95% confidence interval (p-value ≤ 0.05). The statistical significance of response variables (i.e., nanomodifier type, binder type, and rest/healing period) on the healing behavior of PG 58-28 and PG 76-22 asphalt mixtures were determined.
The effect of all the variables on the healing in asphalt binders was found to be significant as observed from Table 2. Nevertheless, the interaction between the variables also displayed statistical significance. Therefore, it can be concluded that the constituents as well as the interaction between them affect the healing potential of modified asphalt binders. In other words, different constituents of the nanomodified binders (modifier type, dosage, rest periods) played an important role in the self-healing capability of asphalt binders and therefore needs to be addressed comprehensively.
Analysis of Variance Results
Statistically significant at 95% confidence level.
Conclusion
This study evaluated the fatigue, healing, rutting, and cracking characteristics of asphalt mixtures prepared with nanoclay- and nanosilica-modified binders. The asphalt mixtures were designed using PG 58-28 and PG 76-22 binders modified at 4% nanomaterial dosage by binder weight, following the UFGS Gradation 3 aggregate specification. Volumetric properties were verified to meet design standards. Fatigue and healing behavior were assessed using Dynamic Modulus (|E*|) tests and cyclic fatigue tests at 800 µε strain amplitude, 10 Hz frequency, and 25°C. Healing potential was quantified by introducing rest periods during fatigue testing and comparing fatigue life. Rutting performance was evaluated using APA at 58°C, while cracking resistance was assessed via the IDEAL-CT test at 25°C. Based on the experimental results and analyses, the following conclusions can be drawn:
Cyclic fatigue testing proved more effective than other fatigue tests for evaluating healing, offering higher sensitivity and repeatability. Its simpler sample preparation and greater responsiveness to damage recovery justify its use in assessing nanomodified asphalt mixtures.
The results from the cyclic fatigue test showed that mixtures with 4% nanoclay modified binders achieved approximately 197% increase in failure cycles than the control mixtures and outperforming nanosilica modified mixtures indicating the nanoclay’s ability of resist fatigue.
Rest periods effectively activated intrinsic healing, with the 20-minute interval producing the greatest extension in fatigue life across all mixtures. While all mixtures exhibited rest-induced recovery, the nanoclay-modified mixtures achieved the highest number of fatigue cycles after the 20-minute rest period, indicating superior damage tolerance and resistance to crack propagation.
The results from the APA test show that nanoclay and nanosilica modified mixtures at 4% dosage improved rutting resistance reduction in rut depth compared with the control mixtures. This result affirms the potential of nanoclay in being a more effective modifier for mitigating permanent deformation.
Cracking resistance evaluated via IDEAL-CT demonstrated that nanoclay and nanosilica modified mixtures exhibited higher CTindex and fracture energy values over the control mixtures indicating the nanomodifiers’ enhanced resistance to cracking.
FlexPAVE simulations showed reduced total damage over 20 years for nanomodified mixtures. Nanoclay with rest periods had the lowest total damage compared with the control mixtures. This confirms the long-term performance benefits of nanoclay modification combined with rest-induced healing in mitigating fatigue-related deterioration in asphalt pavements.
The consistency between the FlexPAVE™ simulation outcomes and laboratory fatigue test results reinforce the reliability of the observed improvements in fatigue resistance and healing behavior. This alignment confirms that rest-induced recovery and nanomodification benefits demonstrated at the material level effectively translate into enhanced long-term pavement performance under realistic field conditions.
Mixtures modified with 4% nanoclay showed the best performance, reducing rut depth by 27% and increasing the CTindex by 125% compared with the control. Introducing a 20-minute rest period enhanced fatigue life by 197%, confirming the healing potential of nanomodified mixtures. Overall, nanoclay significantly improved fatigue resistance, healing efficiency, and structural durability, supporting its use for longer-lasting pavements.
The findings support the use of nanomodifiers, particularly nanoclay, as a viable strategy to enhance the durability and long-term performance of asphalt pavements, contributing to extended service life, reduced maintenance needs, and improved sustainability.
Limitations and Future Work
The present study evaluated the influence of nanoclay and nanosilica on the healing and performance of asphalt mixtures at the mixture scale using cyclic fatigue testing with 10- and 20-minute rest periods. While this work demonstrates that nanomodification enhances fatigue recovery at the mixture scale, other aspects remain open for continued investigation. Future work will incorporate microstructural and molecular-scale characterization (e.g., AFM, molecular dynamics) and evaluate healing across multiple damage levels to establish a more explicit mechanistic link between nanomaterial structure and mixture-level healing response. In addition, upcoming work will include benefit-cost and life-cycle cost analyses to quantify the economic value of nanoclay and nanosilica incorporation and determine optimal dosage for agency implementation.
Footnotes
Authors’ Note
The authors affirm that no AI tools were utilized in the composition of any sections of this manuscript. However, an AI tool (ChatGPT) was employed solely for the purpose of improving grammar for limited number of paragraphs, without altering the scientific content.
Author Contributions
The authors confirm contribution to the paper as follows: study conception and design: Yusuf Mehta, Ayman Ali, Samuel Addai Mensah, and Mohit Chaudhary; data collection and analysis: Samuel Addai Mensah, and Mohit Chaudhary; sample preparation and laboratory testing: Samuel Addai Mensah and Mohit Chaudhary; interpretation of results: Samuel Addai Mensah, Mohit Chaudhary, Ayman Ali, Yusuf Mehta, Ben C. Cox, Mohammed H. Elshaer, and Wade A. Lein; draft manuscript preparation: Samuel Addai Mensah and Mohit Chaudhary. All authors reviewed the results and approved the final version of the manuscript.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study is supported by the Broad Agency Announcement Program and the US Army Engineer Research and Development Center under Contract No. W81EWF11276079. The views and conclusions expressed here are those of the authors and do not necessarily represent the official policies or endorsements of the Broad Agency Announcement Program or ERDC
