Performance Evaluation of Porous Asphalt Mixture Reinforced with Waste Cellulose Acetate Fibers

Abstract

Cellulose acetate fiber (CAF), a typical waste product derived from cigarette filters, has attracted growing attention for its potential reuse in asphalt materials. However, its application in porous asphalt (PA) mixtures remains underexplored. This study investigates the effects of CAF on the performance of asphalt binders and PA-13 mixtures through a series of laboratory tests. The results demonstrate that CAF significantly enhances the high-temperature rheological performance of asphalt binders. A 1% CAF content improved the low-temperature rheological performance of asphalt binder, while a higher CAF content resulted in performance degradation. A fatigue life analysis revealed a parabolic relationship with CAF content with the optimal N_f50 observed at a 1% CAF-a 4.3% increase over the original binder. Compared to 3% lignin fiber (LF)-modified binders, 3% CAF-modified binders exhibited reduced temperature sensitivity in high-temperature performance, at least a 4.6% improvement in low-temperature performance and an 8.4% increase in the fatigue life. As for PA-13 mixtures, the incorporation of CAF progressively improved rutting, moisture and stripping resistance with increasing CAF content, achieving the highest dynamic stability, highest tensile strength ratio and lowest mass loss rate at 5% CAF. The low-temperature performance and fatigue life (S = 0.45) of PA-13 mixtures exhibited a parabolic trend, peaking at 3% CAF. Moreover, the 3% CAF-modified PA-13 mixture demonstrated improved low-temperature performance and fatigue resistance, while exhibiting a slight decrease in high-temperature stability, water resistance and resistance to disintegration. Overall, CAF is a viable alternative to LF for improving the durability and service life of asphalt pavements.

1. Introduction

Porous asphalt (PA) mixtures, recognized as an eco-friendly pavement solution, have emerged as a promising alternative in urban road construction due to their multifunctional characteristics, including superior permeability, acoustic attenuation capacity and urban heat island mitigation potential [1]. Despite these advantages, the coarse aggregate gradation and high void content (typically 18–25%) of PA mixtures negatively impact key engineering properties, such as water stability, aging resistance and mechanical strength [2,3]. Therefore, improving the strength and durability of PA mixtures while maintaining their permeability and environmental adaptability has become a key focus of current research.

Fiber reinforcement technology has proven to be an effective strategy for enhancing the performance of asphalt mixtures. Three primary fiber categories—plant-based fibers, synthetic fibers and mineral fibers—have been extensively utilized in asphalt modification with demonstrated efficacy [4]. A systematic investigation by Kou et al. evaluated the rheological properties of SBS-modified asphalt reinforced with four fiber types: flocculated basalt fibers (FBF), chopped basalt fibers (CBF), lignin fibers (LF) and polyester fibers (PF) [5]. Their findings indicated that all tested fibers improved deformation recovery and fatigue resistance of Styrene–Butadiene–Styrene (SBS)-modified asphalt, with optimal dosages identified as 2% for CBF, 4% for FBF, 4% for LF and 3% for PF. A study by Nazmey et al. revealed superior pavement performance of cellulose fibers over glass wool fibers in PA mixtures, particularly in moisture resistance and raveling resistance [2]. Gupta et al. analyzed the impact of aramid fibers, aramid pulp, glass hybrid fibers and cellulose fibers on the wear resistance and strength of PA mixtures [3]. Their experimental results demonstrated that aramid pulp significantly improved the wear resistance, while glass hybrid fibers excelled in enhancing the indirect tensile strength of PA mixtures. Chen et al. investigated the effects of single- and double-fiber addition technologies on the road performance of the Open Graded Friction Course (OGFC) [6]. The results showed that the double-fiber addition technology provided a balanced improvement in the overall road performance of OGFC mixtures, with the optimal ratio of LF to glass fibers being 0.15:0.15. Similarly, Pang et al. optimized hybrid fiber proportions (LF: ceramic fiber = 1:2), yielding simultaneous improvements in moisture stability and low-temperature cracking resistance of asphalt mixtures [7]. Overall, these studies on fiber reinforcement technology have yielded promising results, providing a strong theoretical foundation and technical support for the application of fiber materials in road engineering.

The pursuit of sustainable and green development in road engineering has become a critical research priority in recent years, driven by growing environmental regulations and concerns over resource scarcity. Conventional fiber materials, such as lignin fibers (LF) derived from raw timber, do not align with the principles of sustainability [8,9]. Likewise, the production of synthetic fibers (e.g., polypropylene and polyester) and mineral fibers (e.g., basalt and glass fibers) involves energy-intensive processes, including complex chemical synthesis and high-temperature melting, leading to substantial carbon emissions and elevated production costs [10]. In this context, recycling and reusing waste fibers have emerged as a sustainable alternative, offering three key environmental benefits: resource conservation, carbon footprint reduction and effective waste management. Regarding the reuse of natural fibers, bamboo fibers were incorporated into PA mixtures, demonstrating a significant enhancement in the permanent deformation resistance of PA mixtures [1,11]. Li et al. recycled waste corn stalks by processing them into fibers, which were then applied to improve the service performance of asphalt and asphalt mixtures [12]. The results indicated that, compared to lignin fibers, the corn stover fiber-modified asphalt and asphalt mixtures exhibited superior road performance. Similarly, a study by Meneses et al. evaluated the impact of sugarcane bagasse fibers on the road performance of OGFC mixtures, finding that their inclusion significantly improved the stiffness and crack resistance of OGFC mixtures [13]. Additionally, Singh et al. explored the feasibility of using sisal, coir and rice straw fibers in asphalt mixtures [14]. The use of waste natural fibers for asphalt mixture reinforcement has demonstrated significant potential in terms of performance enhancement. Furthermore, researchers have actively explored solutions to mitigate the environmental pollution caused by non-biodegradable fibers. For instance, one study highlighted the potential of waste carbon fibers as a sustainable solution for enhancing the induction heating healing of asphalt mixtures [15]. Jin et al. investigated the service performance of asphalt overlay with recycled rubber and tire fabric fiber and demonstrated that the inclusion of tire fabric fibers and recycled rubber markedly improved the rutting resistance, cracking resistance and noise reduction in asphalt mixture [16]. In response to the environmental pressure caused by the large number of waste face masks (FM) during the COVID-19 pandemic, innovative studies have proposed crushing these masks into fiber materials for use in FM-modified asphalt and asphalt mixtures [17,18,19,20,21,22,23]. The findings suggest that FM fiber materials can effectively enhance the high-temperature rheological properties of asphalt [17,18,19]. Moreover, the addition of FM materials improves the Marshall stability, rutting resistance, low-temperature crack resistance and moisture sensitivity of asphalt mixtures [20,21,22,23]. These findings collectively advance sustainable waste management practices while establishing new technological pathways for optimizing road material performance.

Cigarette butts, a prevalent environmental pollutant, contain filters primarily made of cellulose acetate fibers (CAF). Approximately 4.5 trillion cigarette butts are produced annually worldwide, and the natural degradation of CAF filters takes over 10 years [24]. However, the current recycling rates for these filters remain below 1%, leading to severe environmental pollution. Pioneering work by RMIT University researchers established a novel recycling proposal to extract CAF from cigarette waste for modification of asphalt and asphalt mixtures. Their findings revealed that CAF enhanced the thermal stability and mechanical properties of asphalt and asphalt mixtures [25,26,27,28]. Hu et al. further confirmed that waste CAF enhances the physical and rheological properties of asphalt [29]. Additional studies have shown that waste CAF can increase the water stability and crack resistance of asphalt mixtures [30,31]. Ahlawat et al. demonstrated that asphalt mixtures modified with CAF exhibited superior chemical resistance compared to those without CAF [32]. Furthermore, Tatarani et al. demonstrated that waste CAF, as an organic fiber, possesses toughness and strength comparable to LF, making it a promising fiber material for asphalt mixtures [33,34]. CAF, derived from waste cigarette butts, is both economically and environmentally more valuable than LF. In conclusion, current research on the reuse of waste CAF has predominantly focused on dense-graded asphalt mixtures and has achieved some positive findings. The incorporation of cigarette butt material into the dense asphalt mixture can significantly reduce thermal conductivity, while simultaneously improving durability, strength and resistance to rutting when compared to conventional dense asphalt mixture. This innovation not only enhances mechanical strength but also contributes to improved chemical resistance. However, despite the promising performance of CAF in dense-graded asphalt mixtures, there have been no reports addressing its applicability to open-graded porous asphalt (PA) mixtures. Due to the large void structure of PA mixtures, they are more susceptible to issues such as aggregate loss and other related distresses. Therefore, this study aims to fill this research gap by focusing on the performance requirements of PA-13 mixtures.

Based on this, this study proposes the innovative use of waste CAF as a reinforcing material in PA mixtures, aiming to achieve efficient reuse of waste resources. Extensive laboratory experiments will be conducted to investigate the influence of waste CAF on the road performance of asphalt and PA mixtures. The results are anticipated to offer a dual solution for enhancing the performance of porous asphalt and advancing the recycling of solid waste resources.

2. Materials and Methods

2.1. Raw Materials

The asphalt used in this study is high-viscosity asphalt, supplied by Guizhou Qianhe Logistics Co., Ltd. The specific technical parameters are provided in

Table 1. Technical properties of high-viscosity asphalt.

Index	60 °C Viscosity (Pa·s)	Softening Point (°C)	5 °C Ductility (cm)	25 °C Penetration (0.1 mm)
Test value	100,280	81.5	35	52
Standards	≥50,000	≥80	≥30	≥40

, in accordance with the Chinese Technical Specifications for Design and Construction of Porous Asphalt Pavement (JTG/T 3350-03-2020) [35]. The limestone-based mineral filler, fine aggregates and coarse aggregates employed in this study were sourced from Guizhou Expressway Industry Co., Ltd., with all materials complying with the requirements specified in the technical specifications of the Chinese Technical Specification for Construction of Highway Asphalt Pavement (JTG F40-2004) [36].

The fibers used in this study include CAF and lignin fiber (LF). The CAF was derived from self-collected waste cigarette butts through a drying and grinding process. Based on existing studies [26,30], the preparation procedure consisted of several steps: first, the packaging and remaining tobacco were removed to isolate the filter; second, the CAF was dried in an oven at 105 °C for 24 h; and finally, the dried filters were ground into fibers using a high-speed crusher. The preparation process for CAF and the modified asphalt is illustrated in Figure 1.

2.2. Asphalt Preparation and Mix Design

2.2.1. Asphalt Preparation

As illustrated in Figure 1, fiber-modified asphalt was prepared using a high-speed shearing mixer. The CAF content was set at 1%, 3% and 5% by weight of the asphalt, respectively. Additionally, fiber-free asphalt and 3% LF-modified asphalt were used as control groups. The preparation process for the fiber-modified asphalt is as follows: First, the high-viscosity modified asphalt was placed in an oven and heated to a flowing state at 180 °C. Next, the designed amount of fiber was added to the high-viscosity modified asphalt and sheared at a speed of 4500 rad/min for 1 h.

2.2.2. Mix Design

In compliance with the gradation range specified in the Chinese standard JTG/T 3350-03-2020 [35], the PA-13 gradation was selected for the preparation of PA mixtures for subsequent experimental investigations. The corresponding gradation curve for the PA-13 mixture is presented in Figure 2.

According to the Chinese standard JTG/T 3350-03-2020, the recommended asphalt film thickness (h) for PA mixtures is 14 μm. The initial asphalt content was calculated using Equations (1) and (2).

$$A=\left(0.41a+0.41b+0.82c+1.64d+2.87e+6.14f+12.29g+32.77h\right)/{10}^3$$

(1)

$$P_b=h\times A$$

(2)

where A represents the total surface area of the aggregates and a, b, c, d, e, f, g and h correspond to the passing percentages (%) of the 4.75 mm, 2.36 mm, 1.18 mm, 0.6 mm, 0.3 mm, 0.15 mm and 0.075 mm sieves, respectively.

At least four Marshall specimens were prepared based on the initial asphalt content. The air void content and Marshall stability were tested to ensure compliance with the technical requirements specified in JTG/T 3350-03-2020. To further optimize the asphalt content, draindown and scattering tests were conducted with asphalt contents adjusted by ±0.5% and ±1%. The test results were plotted, with the turning point of the scattering curve identified as the minimum asphalt content (OAC₁), and the turning point of the draindown curve defined as the maximum asphalt content (OAC₂). Within the range of OAC₁ to OAC₂, the highest feasible asphalt content that met all performance criteria was selected as the optimal asphalt content, based on the results of the Marshall test. The mix design results of PA-13 mixes containing different fiber content are provided in

Table 2. The PA-13 mixes design results.

Group ID	No Fiber	3% LF	1% CAF	3% CAF	5% CAF
OAC (%)	4.68%	5.00%	4.75%	4.83%	4.95%

.

2.3. Engineering Performance

2.3.1. Asphalt Performance

(1)

Dynamic Shear Rheometer (DSR) Test

The DSR test was performed to measure the complex modulus and phase angle of the asphalt, thereby assessing the influence of CAFs on the high-temperature rheological properties. In accordance with the T06289 method outlined in the Chinese Standard Test Methods for Bitumen and Bituminous Mixtures in Highway Engineering (JTG E20-2011) [37], the test was conducted under a controlled strain mode with a strain level of 12% and a frequency of 10 rad/s. The test temperatures ranged from 58 °C to 82 °C, with a 6 °C interval between each measurement.

(2)

Bending Beam Rheometer (BBR) Test

In accordance with the T06289 method outlined in the Chinese standard JTG E20-2011 [37], the low-temperature rheological properties of the asphalt binder were assessed by measuring the stiffness modulus (S) and creep rate (m) of the asphalt samples. The tests were performed using asphalt binder beams with dimensions of 127 mm × 6.35 mm × 12.7 mm, at three temperature conditions: −6 °C, −12 °C and −18 °C.

(3)

Time Sweep Test

The fatigue performance of the asphalt samples was evaluated using a time sweep test conducted under strain-controlled conditions, with stress applied according to a sinusoidal pattern. The test parameters were set as follows: a strain level of 5.0%, a test temperature of 25 °C and a loading frequency of 10 Hz. According to the AASHTO T315 standard [38], fatigue failure of the asphalt is identified when the complex shear modulus |G*| decreases to 50% of its initial value. The number of loading cycles at this point, N_f50, is defined as the fatigue life. Each fatigue life test was conducted with three replicates, and the average values were reported.

2.3.2. PA-13 Mixture Performance

(1)

Wheel Tracking Test

Referring to the T0719 method specified in the JTG E20-2011 standard [37], the high-temperature rutting resistance of PA-13 mixture specimens was evaluated using a rutting test. The dynamic stability (DS) index was measured at 60 °C, with specimen dimensions of 300 mm × 300 mm × 50 mm. During the test, the temperature was maintained at 60 ± 0.5 °C, and the tire contact pressure was set to 0.7 MPa ± 0.05 MPa.

(2)

Low-temperature Bending Test

Asphalt pavements are susceptible to thermal shrinkage cracks in cold environments due to the accumulation of temperature-induced stresses, which represent one of the primary distresses encountered by asphalt surfaces [39]. In accordance with the JTG E20-2011 standard [37], this study evaluates the low-temperature performance of asphalt mixture using a low-temperature bending test. The rutting specimens were first cut into prismatic beams with dimensions of 250 mm × 30 mm × 35 mm. The test was conducted at −10 ± 0.5 °C with a loading rate of 50 mm/min. The failure strength and strain parameters of the specimens were measured to assess their resistance to low-temperature cracking.

(3)

Freeze–thaw Splitting Test

Under the repeated application of vehicle loads, the negative pressure effect generated during the contact between tires and permeable pavement causes the formation of dynamic water pressure within the voids. This pressure continuously acts within the structure, leading to issues such as aggregate stripping. To assess the water damage resistance of the permeable asphalt mixtures, this study adopted the T0729 method from the JTG E20-2011 standard to conduct freeze–thaw splitting tests [37]. The splitting strength of the PA-13 samples before and after freeze–thaw cycles was measured, and the tensile strength ratio (TSR) was calculated to evaluate the mixture’s resistance to water damage.

(4)

Cantabro Test

The interfacial adhesion between aggregates and asphalt binder in PA mixtures is inherently constrained by their open-graded structure and reduced asphalt content. To evaluate the aggregate–asphalt interfacial bonding performance of CAF-reinforced PA mixtures, the Cantabro test was performed following the standardized procedure specified in Section T0733 of JTG E20-2011 [37]. The mass loss rate was then determined to quantitatively assess the mixture’s abrasion resistance and the bonding strength at the aggregate–asphalt interface.

(5)

Indirect Tensile Fatigue Test

The fatigue characteristics of the PA-13 mixture were investigated through indirect tensile fatigue tests conducted under stress-controlled conditions. A sinusoidal loading pattern with a frequency of 10 Hz was employed to simulate the repeated stress conditions encountered by pavement structures during service. The applied stress levels were systematically established at 0.3, 0.4, 0.5 and 0.6 times the ultimate splitting strength of the PA-13 mixture, which was determined through 15 °C splitting strength tests.

3. Results and Discussion

3.1. Performance Evaluation of Modified Asphalt Binder

3.1.1. High-Temperature Rheological Property

The results of the DSR tests are presented in Figure 3. Both the complex modulus and rutting factors of all asphalt samples exhibited a decreasing trend with increasing test temperature, whereas the phase angle demonstrated an opposite trend. This phenomenon can be attributed to the temperature-dependent viscoelastic properties of asphalt binders, which undergo a transition toward a non-Newtonian fluid state as temperature rises [17]. Correspondingly, the elastic behavior of asphalt gradually weakened, reducing its deformation resistance and recovery ability.

At the same temperature, the high-temperature rheological performance of asphalt binder improved with the increasing CAF content. However, the enhancement effect of CAF diminished with increasing test temperature. Compared to the original asphalt binder, the average complex modulus of asphalt binder modified with 1%, 3% and 5% CAF increased by 26.6%, 72.0% and 133.2%, respectively, while the rutting factor increased by 27.9%, 74.9% and 141.7%, respectively. In contrast, the phase angle decreased by 1.8%, 2.9% and 5.8%, respectively. The incorporation of CAF significantly altered the colloidal structure of asphalt, resulting in an increased proportion of the dispersed phase [40]. The incorporation of CAF significantly altered the colloidal structure of asphalt, resulting in an increased proportion of the dispersed phase [29]. Therefore, the proportion of the elastic component in the asphalt binder increased, enhancing its high-temperature rheological properties [41]. Compared to 3% CAF, the rutting factor of 3% LF-modified asphalt changed by 21.4%, 16.9%, 11.5%, 4.0% and −6.5% as the test temperature increased from 58 °C to 82 °C, indicating that the reinforcing effect of LF decreased with increasing temperature. CAF-modified asphalt displayed lower sensitivity to temperature variations, suggesting its superior temperature susceptibility.

3.1.2. Low-Temperature Rheological Property

The results of the BBR tests are shown in Figure 4. The creep stiffness (S-value) and creep rate (m-value), respectively, reflect the low-temperature flexibility and stress relaxation capacity of the asphalt binder. As shown in Figure 4, the S-value decreased while the m-value correspondingly increased with rising temperature, indicating an improvement in the low-temperature rheological properties and demonstrating temperature sensitivity [42]. Furthermore, at the same test temperature, compared to the original asphalt binder, the S-value of the 1% CAF-modified asphalt binder decreased by an average of 6.5%, while the m-value increased by an average of 3.0%. However, for the 3% and 5% CAF-modified asphalts, the S-value increased by an average of 15.3% and 40.5%, respectively, while the m-value decreased by an average of 4.3% and 15.1%, respectively. These results suggest that 1% CAF can effectively improve the low-temperature flexibility and stress dissipation capacity of the asphalt binder. However, as the CAF content increases, the adsorption and stabilization effects of the fibers within the asphalt binder become more pronounced, leading to a deterioration in the low-temperature rheological properties [43]. Compared to 3% CAF, the 3% LF-modified asphalt exhibited a 4.6% increase in the S-value and a 5.4% decrease in the m-value, indicating a decline in low-temperature performance. These findings suggest that CAF has a less detrimental impact on the low-temperature rheological properties than LF.

3.1.3. Fatigue Property

The results of the time sweep test are shown in Figure 5. As seen in Figure 5a, the complex modulus of all asphalt samples gradually decreased with increasing loading cycles, but the rate of decay varied. For the original asphalt binder, the complex modulus declined slowly in the early stages of loading, showing a distinct steady state until failure. For the 1% CAF-modified asphalt, the complex modulus decreased rapidly with increasing loading cycles, followed by a brief steady period before failure. When the CAF content was increased to 3% and 5%, the rate of decay in the complex modulus accelerated with the number of loading cycles, and no distinct steady state was observed. The addition of CAF increased the initial complex modulus of the asphalt binder, with higher fiber content leading to a greater initial complex modulus [44]. Compared to the 3% CAF-modified binder, the 3% LF-modified asphalt exhibited a higher initial modulus, but its rate of decay was faster.

Figure 5b presents the fatigue life N_f50 of the asphalt binders with varying fiber contents. The results revealed a parabolic trend between CAF content and fatigue life. Specifically, compared to the original asphalt binder, the 1% CAF-modified asphalt exhibited a 4.3% enhancement in N_f50, whereas the 3% and 5% CAF-modified asphalt showed reductions of 2.1% and 33.9%, respectively. This pronounced variation is primarily attributed to differences in fiber dispersion and their consequent effects on binder stiffness. At lower dosages (e.g., 1%), CAF can effectively absorb the asphalt binder and enhance the internal network structure, thereby improving fatigue resistance. However, with increasing CAF content, excessive fiber accumulation may lead to poor dispersion, local agglomeration and elevated stiffness, all of which impair the binder’s ability to relieve stress under repeated loading—ultimately causing a marked decline in fatigue life. Furthermore, a comparative analysis revealed that the 3% LF-modified asphalt exhibited an 8.4% lower N_f50 value than the 3% CAF-modified asphalt. These findings collectively suggest that CAF modification exerts a relatively less detrimental effect on the fatigue performance of asphalt binders compared to LF modification.

3.2. Performance Evaluation of PA-13 Mixture

3.2.1. Influence of CAF on Rutting Resistance

Figure 6 illustrates the results of the wheel tracking test. As the CAF content increased, the DS value improved. Specifically, compared to the unmodified mixture, the DS of the PA-13 mixture with 1%, 3% and 5% CAF increased by 5.9%, 12.8% and 18.7%, respectively. The incorporation of CAF significantly improved the rutting resistance of the asphalt mixture, which was consistent with the findings of Rehman et al. [26,27]. The incorporation of fibers reinforced the skeletal structure of the mixture by occupying the voids between aggregates and providing a micro-reinforcement effect. This enhancement improved the bonding between aggregates, minimized displacement and loosening and increased the mixture’s stability under repeated loading, thereby mitigating permanent deformation. These effects became increasingly pronounced with higher fiber content, leading to a gradual enhancement in DS. Moreover, although the 3% CAF-modified PA-13 mixture exhibited a 3.1% lower DS compared to the 3% LF-modified mixture, it still met the requirements of the Chinese standard JTG/T 3350-03-2020 (no less than 5000 cycles/mm) [35].

3.2.2. Influence of CAF on Low-Temperature Cracking Resistance

The effect of varying CAF contents on the low-temperature cracking performance of PA-13 is shown in Figure 7. Compared to the unmodified mixture, the flexural strength of the 1%, 3% and 5% CAF-modified PA-13 mixtures increased by 5.9%, 12.1% and 19.0%, respectively, while the flexural strain increased by 4.3%, 9.4% and 2.0%, respectively. Overall, the low-temperature performance of the PA-13 specimens initially improved—peaking with the 3% CAF modification—and then declined. This trend may be attributed to an optimal number of fibers enhancing the ductility and toughness of the mixture, thereby enabling it to withstand greater deformation under low-temperature conditions. However, excessive CAF increases overall stiffness and brittleness, weakening the reinforcement effect of fibers on the low-temperature performance [45].

Moreover, compared to the 3% LF-modified mixture, the 3% CAF-modified PA-13 mixture exhibited a 3.2% reduction in flexural strength, while the flexural strain increased by 2.4%, leading to an overall enhancement in low-temperature performance. This difference may result from the poorer dispersion of LFs, which tend to clump and create localized defects that serve as crack initiation points, thereby reducing the overall flexural strength [46]. Additionally, LF’s higher affinity for asphalt results in a stiffer mixture with diminished low-temperature flexibility, as evidenced by the higher flexural strength and lower flexural strain.

3.2.3. Influence of CAF on Moisture Resistance

Figure 8 presents the results of the freeze–thaw splitting strength test. As the CAF content increased, the TSR value of the PA-13 mixture improved. Compared to the unmodified mixture, the 1%, 3% and 5% CAF-modified PA-13 mixtures exhibited increases in the TSR value of 2.0%, 2.8% and 5.4%, respectively. The fibers fill the voids between aggregates, thereby reducing the interconnected pore network and shortening the water penetration pathways. This mechanism slows water intrusion and alleviates the pressure exerted by ice crystal expansion during freeze–thaw cycles, ultimately enhancing the mixture’s resistance to freeze–thaw damage. Moreover, the fibers possess inherent ductility and energy absorption capabilities under load, which enable the mixture to better absorb and dissipate energy during freeze–thaw cycles, thus mitigating the initiation and propagation of cracks [33,47]. As the CAF content increased from 1% to 5%, the internal network structure of the mixture became more robust, leading to a more pronounced crack-blocking effect.

Furthermore, compared to the 3% LF-modified mixture, the 3% CAF-modified PA-13 mixture exhibited a 3.1% reduction in water stability. According to the mixture design results presented in Table 2, the asphalt content in the 3% LF-modified PA-13 mixture is higher than that in the 3% CAF-modified mixture. This difference may be attributed to the porous microstructure of LF, which has a strong asphalt absorption capacity and forms a thick, uniform asphalt film [5,7]. This film effectively reduces water penetration and enhances resistance to water damage.

3.2.4. Influence of CAF on Stripping Resistance

The results of the Cantabro test are shown in Figure 9. As depicted, the mass loss rate (MLR) of PA-13 specimens decreased as the CAF content increased. Specifically, compared to the unmodified mixture, the MLR of the 1%, 3% and 5% CAF-modified PA-13 mixtures decreased by 1.7%, 3.1% and 5.0%, respectively. These results indicate that an increase in CAF content gradually enhances the impact resistance of the mixture while reducing particle loss. This improvement is primarily attributed to CAF reinforcing the adhesion between the asphalt binder and aggregates [31]. Additionally, CAF integrates into the asphalt binder, increasing its overall viscosity, which helps to tightly encapsulate the aggregates and mitigate their detachment [29].

Conversely, the 3% CAF-modified mixture exhibited a 1.2% higher MLR compared to the 3% LF-modified mixture. This increase may be due to the lower asphalt absorption capacity of CAF relative to LF, resulting in a thinner asphalt film around the aggregates. Thinner asphalt films are more prone to detachment under frictional forces during the Cantabro test, exposing aggregates and increasing mass loss. Moreover, the relatively inert surface chemistry of CAF weakens chemical interactions with both asphalt and aggregates, leading to reduced interfacial adhesion compared to LF [30].

3.2.5. Influence of CAF on Fatigue Resistance

Table 3. The results of the 15 °C indirect tensile test.

Group ID	No Fiber	1% CAF	3% CAF	5% CAF	3% LF
Splitting strength (MPa)	0.81	0.89	0.95	1.04	0.98

summarizes the results of the splitting test conducted at 15 °C, indicating that incorporating fibers significantly enhances the splitting strength of the PA-13 mixture. Based on these splitting strength values, the applied load for the indirect tensile fatigue test was determined. The corresponding fatigue test results under different stress ratios are summarized in

Table 4. The results of the 15 °C fatigue test.

Type Sample	The Fatigue Lives (N) Corresponding to Stress Ratio (S)
	0.3	0.4	0.5	0.6
No fiber	8409	2570	1469	656
	8969	2867	1586	730
	7928	2491	1658	699
1% CAF	7360	2905	1807	803
	8109	2709	1527	722
	11,958	2821	1669	775
3% CAF	9614	3302	1780	780
	11,146	2918	1696	819
	9284	3096	1715	855
5% CAF	9585	3235	1623	736
	8455	3039	1578	791
	11,064	2864	1331	637
3% LF	12,056	2900	1507	726
	9923	3376	1846	684
	10,620	3493	1456	735

.

As shown in Table 4, the fatigue life of various PA-13 specimens varied significantly. Furthermore, the fatigue life demonstrated a gradual reduction with increasing stress levels. The equivalent fatigue lives were further analyzed using a double logarithm equation, as shown in Equation (3) [6].

$$\lg N=-b\lg S+a$$

(3)

where parameters a and b denote the equation coefficients, with an increased a and a decreased b contributing to improved fatigue resistance.

Based on the data in Table 4, a linear analysis was conducted using Equation (3). The lgN–lgS curve is depicted in Figure 10a, and the corresponding fatigue equations are presented in

Table 5. Fatigue equation parameters.

Type Sample	a	b	R²
No fiber	2.091	3.480	0.989
1% CAF	2.129	3.464	0.989
3% CAF	2.138	3.526	0.993
5% CAF	2.043	3.695	0.998
3% LF	2.011	3.855	0.997

. As shown in Table 5, the linear correlation coefficients for lgN versus lgS in the various PA-13 mixtures all exceed 0.98, demonstrating that using a double-logarithmic fatigue equation to analyze the fatigue test data is feasible. Furthermore, Figure 10a reveals that parameter a in the fatigue equation initially increased and then decreased with increasing CAF content. The maximum value of a was observed in the 3% CAF-modified PA-13 mixture. Additionally, the increasing trend of parameter b with CAF content suggests that the incorporation of CAF improved the sensitivity of the fatigue performance to stress variations. Compared to the 3% LF-modified mixture, the 3% CAF-modified PA-13 mixture exhibited a larger a value and a smaller b value. Consequently, the fatigue curve for the CAF-modified mixture was positioned lower at low-stress levels and higher at high-stress levels, indicating that its fatigue life was less sensitive to stress variations. These findings imply that CAF was more effective in improving the fatigue life of asphalt mixtures.

Since the tensile stress ratio applied by wheel loads on asphalt pavements is generally no more than 0.45 [7], a further comparative study was conducted on the fatigue life of specimens at S = 0.45, as shown in Figure 10b.

As shown in Figure 10b, compared to the unmodified mixture, the fatigue life of PA-13 increased by 7.9%, 15.8% and 6.5% for 1%, 3% and 5% CAF contents, respectively. This improvement is primarily attributed to the ability of CAF to effectively bridge the aggregates and asphalt, thereby enhancing the structural stability of the PA mixture and improving its capacity for recovery. The three-dimensional fiber network facilitates the dispersion and redistribution of external loads, mitigating stress concentrations and delaying crack initiation [33,45]. Additionally, CAFs impede crack propagation by requiring additional energy for cracks to penetrate the fiber network. However, excessive CAF content may lead to fiber agglomeration, diminishing the reinforcing effect and limiting the overall improvement in fatigue performance.

A comparative analysis indicated that the 3% CAF-modified PA-13 mixture exhibited a 3.2% increase in fatigue life compared to the 3% LF-modified PA-13 mixture. This improvement may be attributed to the better dispersion of CAF within the asphalt mixture, resulting in a stronger structural reinforcement effect. Meanwhile, LF demonstrates a higher asphalt absorption capacity than CAF, imposing greater constraints on the asphalt and limiting the deformation ability—a finding consistent with the low-temperature bending test results [26,27]. Consequently, the incorporation of CAF effectively enhances the flexibility of PA-13, reduces its sensitivity to stress variations and improves fatigue resistance.

4. Conclusions

This study comprehensively investigated the effects of cellulose acetate fiber (CAF) on the performance of asphalt binders and porous asphalt (PA-13) mixtures through a series of laboratory tests and analyses. The key findings are summarized as follows:

(1) The incorporation of CAF significantly improved the high-temperature rheological performance of asphalt binders, with CAF-modified binders exhibiting reduced sensitivity to temperature variations compared to their LF-modified binders.

(2) A 1% CAF content effectively enhanced the low-temperature rheological performance of asphalt binders, whereas higher CAF contents resulted in a deterioration of these properties. Compared to LF, CAF exerted a less detrimental impact on low-temperature properties.

(3) The CAF-modified asphalt binders displayed a parabolic relationship between fiber content and fatigue life, with the optimal performance observed at 1% CAF. The 3% CAF-modified binder demonstrated an 8.4% higher fatigue life than the 3% LF-modified binder, indicating a comparatively lower adverse impact on fatigue performance.

(4) The rutting, moisture and stripping resistance of PA-13 mixtures improved with increasing CAF content, with the best DS, TSR and MLR value observed at 5% CAF. Compared to the unmodified mixture, the addition of 5% CAF enhanced rutting resistance by 18.7%, moisture resistance by 5.5% and stripping resistance by 5.0%, respectively.

(5) The low-temperature performance and fatigue life (S = 0.45) of PA-13 mixtures exhibited a parabolic trend, peaking at 3% CAF. Compared to the unmodified mixture, the addition of 3% CAF resulted in a 9.4% increase in low-temperature performance and a 15.8% increase in fatigue life, respectively.

(6) Compared to the 3% LF-modified mixture, the 3% CAF-modified PA-13 mixture exhibited a 3.1% lower DS, a 2.4% increase in flexural strain, a 3.1% decrease in TSR, a 1.2% improvement in MLR and a 3.2% enhancement in fatigue life.

In conclusion, CAF modification presents a promising strategy for enhancing the performance of asphalt binders and porous asphalt mixtures, making it a viable additive for asphalt pavements. The study identifies 1% CAF as optimal for improving binder properties, particularly rheological and fatigue performance, while 5% CAF is most effective for reinforcing porous asphalt mixtures in terms of rutting resistance, moisture stability and durability. This difference in optimal dosages reflects the distinct mechanisms by which CAF interacts within binder and mixture systems. Therefore, dosage selection in practice should be tailored to the specific application and desired performance outcomes.

Future studies should further explore the PA-13-specific functionalities of CAF-modified mixtures, including drainage performance, noise reduction and skid resistance, to comprehensively evaluate their suitability for porous asphalt applications. In addition, long-term field performance and aging behavior under varying environmental conditions should be investigated to support practical implementation. To support sustainable implementation, a comprehensive life cycle analysis (LCA) is also recommended to quantify the environmental and economic benefits of using recycled CAF in asphalt applications.