Performance Enhancement of Asphalt Mixtures Using Recycled Wind Turbine Blade Fiber

Abstract

To facilitate the sustainable recycling of retired wind turbine blades (RWTBs) and promote the green development of the wind energy sector in China, this study investigates the reuse of crushed RWTBs as composite fiber additives in asphalt mixtures. A systematic optimization of the incorporation process was conducted, and the effects of RWTB fibers on pavement performance were comprehensively evaluated. Using the entropy weight method, the optimal fiber content and particle size were identified as 0.15 wt% and 0.3–1.18 mm, respectively. The experimental results demonstrated that, under optimal conditions, the dynamic stability, low-temperature flexural tensile strain, Marshall stability after water immersion, and freeze-thaw splitting strength of the base asphalt mixture increased by 27.1%, 23.8%, 9.9%, and 8.1%, respectively. Microstructural analyses using SEM and EDS revealed that the reinforcing mechanism of RWTB fibers involves adsorption, bridging, and network formation, which collectively enhance the toughness and elasticity of the asphalt matrix. In addition, a comparative evaluation was performed using the Analytic Hierarchy Process (AHP), incorporating both performance and cost considerations. The comprehensive performance ranking of fiber-modified asphalt mixtures was consistent for both base and SBS-modified asphalt: BF AC-13 > RWTB AC-13 > GF AC-13 > PF AC-13 > unmodified AC-13. Overall, this study confirms the feasibility of high-value reuse of RWTB waste in road engineering and provides practical insights for advancing resource recycling and promoting sustainability within the wind power industry.

1. Introduction

Asphalt pavements are highly susceptible to fatigue cracking and structural failure under repeated traffic loading, leading to progressive degradation over time. To enhance their structural performance, the modification of bitumen—the viscous petroleum-derived binder—has emerged as a promising technological approach. In this context, “bitumen” refers to the binder itself, while “asphalt concrete” (or asphalt mixture) denotes the composite paving material comprising mineral aggregates and bitumen. Among various modifiers, fibers have attracted considerable attention due to their significant reinforcement effects [1,2,3,4,5,6]. Over the past decade, both synthetic and natural fibers—such as glass fiber-reinforced polymers (GFRPs), basalt fibers, polyester fibers, lignin fibers, and polyacrylonitrile fibers—have been extensively studied for asphalt modification [7,8,9,10,11]. These fibers have been shown to significantly improve stiffness, rutting resistance, and fatigue durability across various climatic and traffic conditions. Although some trade-offs, such as reduced low-temperature flexibility, have been reported, optimized combinations of fiber type, dosage, and dimensions have demonstrated promising performance improvements.

Driven by sustainability concerns and the advancement of the circular economy, research has increasingly focused on the reuse of recycled and waste fibers. These materials not only enhance the mechanical properties of asphalt mixtures but also mitigate industrial solid waste generation. For example, waste tire textile fibers (WTTFs) exhibit notable reinforcement potential, although challenges such as rubber contamination and separation complexity persist [12]. Other recycled materials—including cigarette filters [13], rock wool fibers [14], and textile waste fibers [15]—have also been applied in asphalt mixtures, yielding varying degrees of mechanical and environmental benefits and reflecting the broader shift toward sustainable pavement engineering.

Meanwhile, the growing scale of decommissioned wind turbine blades (WTBs) presents a critical waste management challenge worldwide. WTBs are primarily composed of thermoset resin and glass fibers, featuring highly cross-linked molecular structures that are notoriously difficult to recycle. Global projections estimate that over 2.7 million tonnes of retired wind turbine blades (RWTBs) will be generated by 2055 [16,17,18,19], making their high-value reuse an urgent engineering concern.

In the field of building construction, the multifunctionality of RWTB materials has already been preliminarily validated. Sorte et al. [20] applied shredded RWTB materials in building insulation systems and observed favorable acoustic absorption properties, with thermal performance significantly influenced by fiber length. Similarly, Baturkin et al. [21] incorporated RWTB-GFRP waste into lightweight concrete as aggregate, achieving notable reductions in carbon emissions despite slight strength losses. These studies demonstrate the potential of RWTB materials for interdisciplinary reuse.

In road engineering, RWTB fibers have also drawn increasing attention as asphalt modifiers. Yu et al. [22] explored the feasibility of using RWTB powder and fibers as lightweight fillers and reinforcement in asphalt mixtures, emphasizing their circular resource potential. Wang et al. [23] incorporated RWTB fibers into deicing asphalt mixtures and optimized fiber content (0.28–0.29 wt%) using a multi-objective decision-making method, significantly enhancing cracking resistance and structural strength. Li et al. [24] developed a silane- and SBR-treated RWTB composite modifier, demonstrating excellent performance in moisture resistance, rutting resistance, and freeze–thaw durability. Lan et al. [25] co-incorporated RWTB powder and chopped fibers into asphalt, improving interfacial compatibility with bitumen and further enhancing cracking resistance, rutting resistance, and moisture stability. Nie et al. [26] investigated the mechanical behavior of recycled RWTB glass fibers in SBS-modified asphalt, showing substantial improvements in elasticity and toughness.

Although these studies have preliminarily confirmed the reinforcement potential of RWTB fibers in asphalt, the current literature remains largely limited to proof-of-concept evaluations. There is a lack of standardized preparation protocols, comprehensive performance comparisons with conventional fibers, systematic optimization of particle size and dosage, and in-depth exploration of reinforcement mechanisms at the microstructural level.

To address these research gaps, this study focuses on evaluating the performance of RWTB fibers in asphalt mixtures. A 9-factor, 4-level orthogonal experimental design combined with gray relational analysis is employed to identify the optimal particle size and content. Microstructural characteristics and reinforcement mechanisms are examined using scanning electron microscopy (SEM) and Fourier transform infrared spectroscopy (FTIR). Comparative performance evaluations are conducted against conventional fibers including glass (GF), polyester (PF), and basalt (BF), and a multi-criteria assessment framework incorporating the Analytic Hierarchy Process (AHP) is developed to rank fiber types based on both road performance and cost metrics. This forms an integrated research model centered on “performance optimization–micromechanism analysis–multi-criteria decision-making.”

This study not only offers a systematic approach to the high-value reuse of RWTB waste in road engineering but also contributes theoretical foundations and practical strategies for the development of green, low-carbon, and sustainable pavement materials. By advancing the circular application of wind turbine blade waste in asphalt pavements, the research findings are expected to support dual objectives of waste reduction and carbon emission control, in alignment with the core sustainability goals of modern infrastructure development.

2. Materials and Methods

2.1. Materials

2.1.1. Bitumen

This study used 60/70 penetration-grade bitumen and I-D grade SBS-modified bitumen produced by Shell Co., Ltd., Shanghai, China. The detailed properties of the asphalt materials are presented in

Table 1. Technical specifications for 60/70 penetration-grade bitumen and SBS modified bitumen.

Index	Pen 60/70 Base Bitumen		SBS Modified Bitumen		Test Method
	Test Value	Specification	Test Value	Specification
Penetration (25 °C)/0.1 mm	68.3	60~80	52.8	40~55	T0604
Softening point/°C	48.0	≥45	85.0	≥75	T0606
Ductility (15 °C)/cm	>100	≥100	—	—	T0605
Ductility (5 °C)/cm	—	—	32.2	≥25
Flash point/°C	297	≥260	285	≥230	T0611
Density (15 °C)/g·cm−3	1.033	―	1.039	—	T0603

, as determined according to the Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering (JTG E20-2011) [27].

2.1.2. Asphalt Mixture

The aggregates used in this study included coarse and fine limestone aggregates and manufactured limestone sand. The coarse aggregates were classified into three size ranges, 9.5–13.2 mm, 4.75–9.5 mm, and 2.36–4.75 mm, while the fine aggregate had a particle size of 0–2.36 mm. The mineral filler was a finished product manufactured by Guangxi Senle Building Materials Trading Co., Ltd., Liuzhou, China. The results are summarized in

Table 2. Basic technical indexes of coarse and fine aggregates.

Coarse Aggregate
Index		Unit	Technical requirement	Test result	Test method
Crushed stone value (9.5~13.2 mm)		%	≤28	12.2	T0316-2005
Apparent relative density	9.5~13.2 mm	t/m3	≥2.5	2.725	T0304-2005
	4.75~9.5 mm	t/m3		2.726
	2.36~4.75 mm	t/m3		2.716
Water absorption	9.5~13.2 mm	%	≤3.0	0.35	T0304-2005
	4.75~9.5 mm	%		0.53
	2.36~4.75 mm	%		1.02
Bulk relative density	9.5~13.2 mm	t/m3	—	2.699	T0304-2005
	4.75~9.5 mm	t/m3		2.688
	2.36~4.75 mm	t/m3		2.643
Flakiness index	9.5~13.2 mm	%	≤15	9.6	T0312-2005
	4.75~9.5 mm	%		10.7
	2.36~4.75 mm	%		8.2
Passing 0.075 mm sieve by washing method	9.5~13.2 mm	%	≤1	0.2	T0302-2005
	4.75~9.5 mm	%		0.2
	2.36~4.75 mm	%		0.4
Fine aggregate
Apparent relative density		t/m3	≥2.50	2.700	T0328-2005
Bulk relative density		t/m3	—	2.698
Passing 0.075 mm sieve by washing method		%	≤3	2.6	T0327-2005
Sand equivalent		%	≥60	70	T0334-2005

and

Table 3. Basic technical indicators of mineral powder.

Index		Unit	Technical Requirement	Test Result	Test Method
Apparent density		t/m3	≥2.50	2.691	T0352-2000
Particle size range	<0.6 mm	%	100	100	T0351-2000
	<0.15 mm	%	90~100	99.8
	<0.075 mm	%	75~100	95.5
Water affinity coefficient		—	<1 (preferably <0.8)	0.6	T0353-2000
Plasticity index		—	≤4	2.4	T0354-2000

.

The aggregate gradation of the RWTB fiber-reinforced asphalt mixtures was designed using the AC-13 gradation in accordance with the Technical Specifications for Construction of Highway Asphalt Pavements (JTG F40-2004) [28], as shown in

Table 4. Mineral aggregate design gradation of AC-13 asphalt mixture.

Sieve Size/mm	16	13.2	9.5	4.75	2.36	1.18	0.6	0.3	0.15	0.075
Synthetic grade/%	100.0	92.8	72.3	39.0	24.9	17.9	12.0	8.7	7.0	5.0
Upper limit of grading/%	100	100	85	68	50	38	28	20	15	8
Lower Grading Limit/%	100	90	68	38	24	15	10	7	5	4

. The corresponding Marshall test results are presented in

Table 5. Optimal bitumen-to-stone ratio Marshall test results for AC-13 asphalt mixture.

	Gross Volume Relative Density	Theoretical Maximum Relative Density	Void Ratio/%	VMA/%	VFA/%	Degree of Stability/kN	Stream Value/mm
Result	2.428	2.512	4.8	13.7	68.3	12.80	2.4
Specification	—	—	3–6	>13	65–75	>8	2–4

, from which the optimal asphalt content was determined to be 4.8%.

2.1.3. RWTB Fibers

The RWTB fibers used in this study were obtained from mechanically crushed raw materials and supplied by an industrial materials company in Hunan. The processes and characterization process are detailed as follows. Approximately 400 g of fibers were subjected to a particle distribution test by using a ZBSX-89 shaking screen machine for 30 min. Fibers were separated into nine ranges by determination of the substance retained on standard sieves. Each part was re-sieved for 15 min for precision and retention, and the part on top of the sieve was gathered for determination. The outcome from re-sieving is depicted in Figure 1 and

Table 6. Classification and screening of WFB raw materials.

Particle Size	Percentage/%	Appearance Description
>9.5 mm	4.7	Predominantly flake resin and impurities
9.5–4.75 mm	14.9	More flaky resin and balsa with attached fibers
4.75–2.36 mm	17.4	Strips of resin, fibers, increased particles
2.36–1.18 mm	12.5	Strips of resin with more small particles
1.18–0.6 mm	18.1	Striped fibers agglomerated with resin, small particle size
0.6–0.3 mm	10.9	Fibers agglomerated and attached around striated fibers, fewer particles
0.3–0.15 mm	6.1	White appearance, fluffy volume, fiber agglomeration and mixed with resin
0.15–0.075 mm	7.3	Short fibers agglomerated with resin, soft to the touch, dusty
<0.075 mm	8.1	More fine dust, fiber and dust adsorption

.

It is worth noting that prior research conducted by our group demonstrated that the use of silane-treated RWTB fibers in asphalt modification resulted in a slight increase in penetration, a nearly unchanged softening point, significantly improved ductility, and reduced viscosity. Overall, the low-temperature performance of the asphalt binder was moderately enhanced. However, the silane treatment process is relatively complex, time-consuming, and costly. Although it improves low-temperature properties, the overall cost-effectiveness was found to be low [29]. Therefore, in this study, the RWTB fibers were used without any surface treatment.

2.1.4. Comparison of Multiple Fibers

Glass fiber (GF), polyester fiber (PF), and basalt fiber (BF), all supplied by an industrial materials company in Hunan, were selected as typical reference materials to evaluate the performance improvement of RWTB fibers, as they have been widely reported to enhance the pavement performance of asphalt mixtures [30].

Table 7. Technical specifications of RWTB fiber and comparison fibers.

Fiber Types	Fiber Length/ mm	Fiber Diameter/ μm	Density/(g/cm3)	Safety
RWTB fiber	5~8	15	1.58	non-toxic
GF	6	13	2.51	non-toxic
PF	6	13.4	1.36	non-toxic
BF	6	17	2.65	non-toxic

provides a comparison between the technical specifications of RWTB fiber and those of the three conventional fibers. To avoid ambiguity, it should be clarified that “0.3–1.18 mm” refers to the particle size range obtained from sieve analysis, whereas “5–8 mm” refers to the predominant fiber length measured separately by optical microscopy. Asphalt mixes corresponding to each fiber type were also prepared using their corresponding optimum fiber contents and asphalt-aggregate ratios and compared for pavement behavior [31].

2.2. Experiments

To evaluate basic properties and pavement behavior of modified asphalt mixes with RWTB materials, laboratory experiments were carried out. The experiments included testing high-temperature rutting resistance, cracking resistance at low temperature, and moisture damage resistance. These experiments were carried out based on Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering (JTG E20-2011).

For each performance test, a total of 21 specimens were prepared, including replicates. Among them, 12 specimens were used for analyzing the optimal RWTB fiber dosage (four dosage levels with three replicates each), and 9 specimens were used for comparative analysis with conventional fibers (GF, PF, and BF, three replicates each).

2.2.1. Marshall Stability Test

As specified in the compaction method specified in T0702 of JTG E42-2005 [32], Marshall specimens were prepared as follows: Aggregates, asphalt, and mineral powder were added sequentially in the designed proportions and thoroughly mixed. The mixture was then quickly placed into a mold and compacted on both sides with 75 blows using a Marshall Electric Compactor, SYD-0702 (Shanghai Changji Geological Instrument Co., Ltd., Shanghai, China). Standard Marshall specimen molds with a diameter of 101.6 mm and a height of 63.5 mm ± 1.3 mm were measured using a vernier caliper. After cooling at room temperature for more than 24 h, the specimens were demolded and used for the Marshall stability test. Marshall stability is a measure of the maximum load carrying capacity of an asphalt mixture at high temperature and is a measure of its resistance to deformation at high temperature. Marshall specimens and also the upper and lower platens of a test machine were conditioned before testing in a 60 °C water bath for 30 min. Subsequently, we tested them using a Microcomputer-Controlled Marshall Stability Tester, MDW-50 (Hangzhou Xingao Technology Co., Ltd., Hangzhou, China), with a loading rate of 50 mm/min until failure. Stability was recorded automatically.

2.2.2. Rutting Test

In compliance with T0703-2011 (wheel-tracking method) of JTG E42-2005, rutting tests were performed on slab specimens with dimensions of 300 mm × 300 mm × 50 mm. After molding, the specimens were cured for at least 12 h at room temperature and then conditioned at 60 ± 1 °C for 5–12 h. During testing, the temperature was maintained at 60 ± 0.5 °C, and the contact pressure between the specimen and the test wheel was 0.7 ± 0.05 MPa. Tests were conducted using an Automatic Rutting Tester, HYCZ-1 (Beijing Aerospace & Aviation Measurement and Control Technology Institute, Beijing, China). Each specimen was placed centrally on the wheel with the rolling direction aligned parallel to the compaction direction. The test was carried out for 1 h or until a rut depth of 25 mm was reached, whichever occurred first. The wheel load reciprocating speed was set to 42 cycles/min with a travel distance of 230 mm per cycle. The dynamic stability (DS) was calculated according to JTG E20-2011 as follows:

$$DS=\frac{\left(t_2-t_1\right)\times N}{d_2-d_1}\times C_1\times C_2$$

(1)

where DS is the dynamic stability (cycles/mm), N is the wheel load reciprocating speed (42 cycles/min), C₁ is the instrument type coefficient, C₂ is the specimen coefficient (1.0 when specimen width is 300 mm), and d₆₀ and d₄₅ are rut depths at 60 and 45 min, respectively. If rut depth reached 25 mm before 60 min, the test was stopped, and the deformation values were recorded accordingly.

2.2.3. Low-Temperature Bending Beam Test

As required in T0715 of JTG E42-2005, beam specimens with dimensions of 250 mm × 30 mm × 35 mm were conditioned at −10 °C for at least 45 min prior to testing. Testing was conducted on a Microcomputer-Controlled Universal Testing Machine, WDW-100E (Jinan Shidai Shijin Testing Machine Co., Ltd., Jinan, China), using three-point bending with a 200 mm span and loading rate of 50 mm/min.

2.2.4. Immersion Marshall Test

According to T0709 in JTG E42-2005, the Marshall specimens were divided into two groups. One group was immersed in a 60 °C water bath for 30 min and the other for 48 h, before testing on a Microcomputer-Controlled Marshall Stability Tester, MDW-50 (Hangzhou Xingao Technology Co., Ltd., Hangzhou, China). The ratio of the stability values from the two groups was calculated as the immersion residual stability, which reflects the moisture resistance of the asphalt mixture.

2.2.5. Freeze–Thaw Splitting Test

Following the requirements of T0729-2000 in JTG E42-2005, standard Marshall specimens were prepared by compacting each side 50 times. One was conditioned at room temperature and thereafter put in a 25 °C water bath for 2 h and tested. Another was vacuum-saturated for 15 min and conditioned at −18 °C for 16 h and thereafter conditioned at a 60 °C water bath for 24 h and put in a 25 °C water bath for 2 h. Tests were conducted on a Microcomputer-Controlled Universal Testing Machine, WDW-100E (Jinan Shidai Shijin Testing Machine Co., Ltd., Jinan, China), to determine the freeze–thaw splitting strength ratio. The freeze–thaw splitting strength ratio from these two was utilized to estimate moisture resistance from the asphalt mixture.

2.2.6. Scanning Electron Microscopy (SEM) Test

RWTB fibers were placed in the chamber of a Scanning Electron Microscope, Sigma 300 (Carl Zeiss AG, Oberkochen, Germany), after appropriate surface preparation. Imaging was carried out at selected magnifications using an appropriate scanning rate. An EDS spectrometer (Oxford Instruments, Abingdon, UK) was also used on selected areas to determine the elemental composition.

2.2.7. Fourier Transform Infrared Spectroscopy (FTIR) Test

A Fourier Transform Infrared Spectrometer, Nicolet 6700 (Thermo Fisher Scientific Inc., Waltham, MA, USA) (FTIR), was used for characterization of infrared spectra of both RWTB fibers and asphalt binders. For RWTB fibers, the KBr pellet transmission method was employed: the fibers were first dried and ground with spectroscopically pure potassium bromide (KBr), then pressed with a mold to prepare pellets for recording the infrared spectra. For base asphalt and RWTB composite asphalt binders, the attenuated total reflection (ATR) mode was applied, enabling direct measurement of thin asphalt films at room temperature without additional sample preparation. All ATR-FTIR tests were conducted at room temperature on base asphalt, RWTB composite fibers, and RWTB composite asphalt, as shown in Figure 2, with a scanning range of 4000–600 cm⁻¹, a resolution of 4 cm⁻¹, and 32 scans.

2.3. Preparation Process of RWTB Fiber-Modified Asphalt

To optimize the preparation process of RWTB fiber-modified asphalt mixtures, four key parameters were considered: fiber preheating temperature, aggregate heating temperature, mixing temperature, and mixing time [33]. Each parameter was set at three levels: 120 °C, 150 °C, and 180 °C for fiber heating; 160 °C, 170 °C, and 180 °C for aggregate heating; 160 °C, 170 °C, and 180 °C for mixing temperature; and 60 s, 70 s, and 80 s for mixing time. An L9 (3^4) orthogonal test design was employed to conduct Marshall tests. Based on the results, the optimal preparation process for RWTB fiber-modified asphalt mixtures was determined, as illustrated in Figure 3.

2.4. Optimal Particle Size Range

To determine the optimal particle size range for RWTB fibers, a fiber content of 0.2% and the previously determined optimal asphalt content were adopted. Specimens were prepared using the mixing procedure described in Section 2.1. The results are summarized in

Table 8. AC-13 Marshall test results with single-size RWTB fiber additives.

WFB Particle Size on Screen/mm	Oil-Rock Ratio/%	Gross Volume Relative Density	Theoretical Maximum Relative Density	Void Ratio/%	VMA/%	VFA/%	Degree of Stability/KN	Stream Value/mm
Fiber-free	4.80	2.428	2.512	3.4	13.7	68.3	12.8	2.4
Bottle	4.90	2.421	2.471	1.9	13.2	82.7	13.00	3.0
0.075	4.85	2.412	2.508	3.6	13.5	69.6	13.74	2.5
0.15	4.85	2.409	2.502	3.5	13.6	70.5	12.84	3.0
0.3	4.85	2.414	2.499	3.2	13.5	72.5	14.37	2.9
0.6	4.85	2.424	2.495	2.7	13.1	76.4	13.99	2.6
1.18	4.85	2.426	2.498	2.3	12.6	78.8	14.49	2.7
2.36	4.80	2.428	2.506	2.9	12.9	74.3	13.64	2.9
4.75	4.80	2.415	2.508	3.5	13.4	70.2	13.69	3.3
9.5	4.80	2.408	2.499	3.4	13.7	71.2	13.13	2.4
Specification	—	—	—	3–6	>13	65–75	>8	2–4

and

Table 9. AC-13 Marshall test results with multiple -size RWTB fiber additives.

WFB Particle Size on Screen/mm	Oil-Rock Ratio/%	Gross Volume Relative Density	Theoretical Maximum Relative Density	Void Ratio/%	VMA/%	VFA/%	Degree of Stability/KN	Stream Value/mm
Fiber-free	4.80	2.436	2.513	3.7	13.6	70.7	12.66	2.4
Bottle~0.6	4.90	2.431	2.498	3.1	13.5	78.2	13.43	2.8
1.18~9.5	4.85	2.430	2.512	4.4	13.6	62.8	13.82	2.4
0~0.15	4.90	2.434	2.499	3.0	13.5	75.5	13.60	2.1
0.3~1.18	4.85	2.424	2.515	3.6	13.8	68.8	14.63	2.5
2.36~9.5	4.85	2.432	2.512	4.0	13.2	65.0	12.88	2.8
Unscreened	4.90	2.437	2.505	2.7	13.3	74.8	12.93	2.2
Specification	—	—	—	3–6	>13	65–75	>8	2–4

. As shown in Table 9, all AC-13 mixtures with single-size fibers met the Marshall flow requirements, although some mixtures exceeded the limits for air voids, saturation, or void ratio. Marshall stability increased with particle size up to a point and then declined, with 0.3 mm, 0.6 mm, and 1.18 mm showing the highest values.

For the eight mixtures containing blended fiber sizes, all key performance indicators fell within specification limits. Among them, fibers in the range of 0.3 to 1.18 mm produced the highest Marshall stability, with a 15.6% improvement. Thus, 0.3–1.18 mm was identified as the optimal particle size range for RWTB fibers.

2.5. Methods

This study seeks to optimize the preparation process of RWTB fiber-modified asphalt mixtures through orthogonal testing of key parameters, including fiber heating temperature, aggregate heating temperature, mixing temperature, and mixing time. The specific preparation process is illustrated in Figure 4. The optimal particle size for RWTB fibers is determined using Marshall tests. Comprehensive performance evaluations are performed using immersion Marshall tests, rutting tests, freeze–thaw splitting tests, and low-temperature bending tests to systematically assess the mechanical properties and road performance of the asphalt mixtures. To explore the functional mechanisms of RWTB fibers at the molecular level, microscopic techniques such as SEM and FTIR are employed. Furthermore, a macro-scale performance comparison is conducted against conventional fibers, including glass, polyester, and basalt fibers, to benchmark the enhancement effects of RWTB fibers. The ultimate objective of this study is to improve the durability, sustainability, and environmental performance of asphalt materials by innovatively incorporating recycled RWTB fibers into their design.

3. Results and Discussion

Based on the preparation flowchart and optimal particle size range determined from the above experiments, asphalt mixtures with varying RWTB fiber contents were prepared. By comparing the pavement performance of these mixtures, the optimal RWTB fiber content was identified.

3.1. High-Temperature Stability

As shown in Figure 5, the dynamic stability of asphalt mixtures increased with the addition of RWTB fibers, reaching a peak at a fiber content of 0.15%, and then declined at higher dosages. Specifically, the maximum dynamic stability of the unmodified base asphalt mixture was 1211 cycles/mm, while the RWTB fiber-modified mixture at 0.15% content achieved 1539 cycles/mm, representing an increase of approximately 27.1%. For the SBS-modified asphalt mixture, the maximum dynamic stability reached 14,916 cycles/mm at 0.15% fiber content, which is about 18.3% higher than that of the fiber-free SBS-modified mixture.

The observed improvements may be attributed to the formation of a more stable internal structure, possibly due to the physical reinforcement provided by well-dispersed RWTB fibers at optimal content. These fibers are assumed to act as a three-dimensional reinforcing network, improving resistance to deformation under loading. However, when the fiber content exceeds 0.15%, a decrease in dynamic stability is observed, which may be caused by fiber agglomeration or poor dispersion, leading to weakened structural integrity [34].

3.2. Low-Temperature Cracking Resistance

The bending test results at low temperature are given in Figure 6. As plotted in Figure 6a, flexural tensile deformation first increases and then decreases by increasing content of RWTB fiber, reaching its maximum value at 0.15%. Flexural tensile deformation of base asphalt and SBS-modified mixes increases by 23.8% and 12.9%, respectively, compared with their corresponding unreinforced mixes at 0.15% fiber content by weight.

The low-temperature stiffness modulus, as shown in Figure 6b, initially decreases with fiber addition and reaches its minimum at 0.15% fiber content, indicating improved flexibility and cracking resistance. These results suggest that the incorporation of RWTB fibers can effectively enhance the low-temperature performance of asphalt mixtures. This enhancement is presumed to arise from the formation of a fiber network structure that helps to redistribute thermal stress and delay crack propagation, a mechanism similarly reported by Liu et al. [35].

3.3. Water Stability Performance

The Marshall immersion test results are shown in Figure 7. It can be observed from Figure 7a that Marshall immersibility and stability of base and SBS-modified asphalt mixes improved after fiber additions using RWTB and achieved their optimum levels when fiber content was 0.2%. Marshall immersibility and stability for fiber-free mixes improved by 11.7% and 2.3%, respectively. It can also be noted from Figure 7b that freeze–thaw splitting tensile strength ratios initially improved and degraded after increasing fiber content and attained an optimum level at 0.15% fiber content. For optimum content, strength ratios improved by 8.1% and 2.5% for base and SBS-modified asphalt mixes, respectively.

These results are consistent with the findings reported by Xie et al. [36], which suggest that the improved moisture stability of fiber-reinforced asphalt mixtures under freeze–thaw conditions may be attributed to the formation of a fiber network. This network structure can reduce moisture migration and increase the thickness of the asphalt film. However, since SBS modification alone significantly enhances moisture resistance, the additional benefit provided by RWTB fibers in SBS-modified mixtures appears to be limited.

3.4. SEM Analysis

SEM tests were performed on particles with sizes ranging from 0.3 to 0.15 mm obtained from crushed RWTB. The SEM images observed at different magnifications are shown in Figure 8. As seen in Figure 8a, the crushed material in this particle size range contains a high proportion of fibers. These fibers exhibit a regular cylindrical shape, with uniform thickness and an interwoven arrangement, having diameters of approximately 10–15 µm and lengths exceeding 1 mm. Fibers that are excessively long or thin tend to agglomerate during the mixing and modification process, resulting in uneven blending. Conversely, if the aspect ratio of the fibers is too small, their reinforcing effect on the mechanical properties of asphalt becomes negligible. The fiber surfaces are not smooth and are embedded with multiple resin granules of varying sizes, as shown in Figure 8d. In addition, as shown in Figure 8c, the crushed material also contains block-shaped solid particles in addition to fibers. Although the current study did not directly characterize fiber dispersion within the asphalt matrix, the previous literature [31] has suggested that well-dispersed fibers with similar morphology can enhance mixture cohesion by forming a reinforcing network. Furthermore, their high modulus and tensile strength may help fortify the asphalt mixture, restrain crack propagation and aggregate slippage, and improve thermal stability. These assumptions are consistent with findings reported in earlier studies.

The EDS spectral analyses were conducted on the blocky solid material in Figure 9a and the fibers in Figure 9b, with the results shown in and Figure 8. As indicated by Figure 8, the blocky solid material is mainly composed of C and O, with C accounting for as much as 41.52% by mass and O accounting for 32.98%. Based on the primary materials and composition of retired wind turbine blades, it was identified as resin particles. The EDS spectrum of the fibers shows that their main components are Si and O, with Si accounting for 30.53% by mass and O for 38.94%, indicating that the primary component is SiO₂ and that the fibers are glass fibers. In EDS analysis, the presence of aluminum, silicon, magnesium, and calcium reflects their use in the fibers to enhance the material’s mechanical properties [37].

3.5. FTIR Analysis

The results are shown in Figure 10. The spectrum of RWTB composite fibers exhibits a broad and prominent peak at ~3450 cm⁻¹, attributed to the O–H stretching of hydrogen-bonded hydroxyl groups/adsorbed water; a peak at 2960 cm⁻¹ corresponding to CH₃; a peak at 1722 cm⁻¹ assigned to the C = O stretching vibration of carbonyl groups (originating from the fiber resin/sizing layer); and a strong, broad absorption at ~1090 cm⁻¹ associated with the asymmetric stretching of Si–O–Si. Combined with the aforementioned EDS results, it can be confirmed that RWTB composite fibers are primarily composed of resin and glass fibers.

When comparing RWTB fiber-modified asphalt with base asphalt, the main absorption peak positions are essentially consistent, with only differences in intensity. This indicates that, under the conditions of this study, no new significant characteristic absorptions appear, and the chemical backbone of the asphalt remains intact. The modification mechanism is dominated by physical adsorption and entanglement, although weak interactions such as hydrogen bonding cannot be ruled out. Due to their large specific surface area, RWTB composite fibers can adsorb part of the light oil fraction, increasing the relative content of asphaltenes/colloidal phase in the system, thereby improving the high-temperature performance [38].

3.6. Determining the Optimal Dosage

Using IBM SPSS Statistics 28.0 (IBM Corp., Armonk, NY, USA) software, the relationship between RWTB fiber content and various pavement performance indicators was analyzed through gray relational analysis [29]. The results, shown in

Table 10. Gray correlation analysis results between different dosage and road performance.

Asphalt Types	Evaluation Items	Peak Dosage/%	Correlation Degree	Ranking
Base asphalt	Bending tensile strain	0.15	0.673	1
	Dynamic stability	0.15	0.661	2
	Residual stability	0.20	0.641	3
	Freeze–thaw splitting tensile strength	0.15	0.635	4
	Elastic modulus	0.15	0.583	5
SBS modified asphalt	Dynamic stability	0.15	0.681	1
	Bending tensile strain	0.15	0.673	2
	Residual stability	0.20	0.657	3
	Freeze–thaw splitting tensile strength	0.15	0.655	4
	Elastic modulus	0.15	0.633	5

, indicate that both the base asphalt composite fiber asphalt mixture and the SBS-modified asphalt mixture exhibit the highest correlation with flexural tensile strain and dynamic stability, both of which reach their peak values at an RWTB fiber content of 0.15%. At a fiber content of 0.2%, the residual stability performs optimally, ranking third in correlation with fiber content. To further determine the optimal dosage of RWTB fibers, entropy weight analysis was conducted [39]. The results, presented in

Table 11. Evaluation and ranking of road performance scores of asphalt mixtures with different dosages.

Asphalt Types	RWTB Fiber Content/ %	Overall Evaluation Score	Ranking
Base asphalt	0.15	0.988	1
	0.2	0.712	2
	0.1	0.347	3
	0.0	0.000	4
SBS modified asphalt	0.15	0.981	1
	0.2	0.732	2
	0.1	0.315	3
	0.0	0.000	4

, show that the comprehensive evaluation rankings of RWTB fiber content are the same for both base asphalt mixtures and modified asphalt mixtures: 0.15% > 0.2% > 0.1% > 0%. In summary, the optimal dosage of RWTB fibers is determined to be 0.15%.

4. Evaluation of the Comprehensive Performance of Four Types of Fibers Based on the AHP Method

4.1. Comparative Analysis of Pavement Performance and Cost of Four Types of Fibers

Table 12. Optimal dosage reference value.

Fiber Types	Optimal Dosage/%
	High-Temperature Stability	Low-Temperature Cracking Resistance	Water Stability
RWTB fiber	0.15	0.15	0.15
GF	0.3	0.2	0.4
PF	0.1	0.1	0.1
BF	0.5	0.5	0.3

summarizes the reference optimal dosages for the four fibers evaluated in this study; because these values depend on the test protocols and boundary conditions [31], they should be regarded as study-specific rather than universal. AC-13 mixtures with each fiber were then prepared at their respective optimum binder–aggregate ratio and fiber content, and their pavement performance was compared to assess the modification effectiveness; the results are shown in Figure 11.

It can be found from Figure 11a that for base asphalt AC-13, adding the optimal dosage of RWTB fiber, GF, PF, and BF increased dynamic stability by 27.1%, 26.2%, 45.7%, and 107.8%, respectively; for SBS-modified AC-13, the corresponding increases were 18.3%, 16.1%, 26.5%, and 28.0%. The similar improvements achieved by RWTB fiber and GF indicate that the RWTB fibers in the 0.3–1.18 mm size range consist largely of relatively pure glass fibers, and AC-13 with RWTB fibers exhibits slightly better high-temperature stability than with glass fibers. Overall, the enhancement in high-temperature stability ranks as BF > PF > RWTB fiber > GF. It can be identified from Figure 11b,c that additions of fibers except PF enhanced cracking resistance at service temperature. The highest improvement was found from BF and successive improvement was from RWTB fiber, GF fiber, and PF fiber which decreased flexural tensile strain.

Figure 11d shows enhanced Marshall residual stability for every fiber-modified mix. Residual stability was enhanced by 11.7%, 10.5%, 14.5%, and 17.5% for RWTB fibers, GF, PF, and BF modification on base asphalt mixes. Enhancements for SBS-modified mixes also followed an identical trend. Improvement ranking was BF > PF > RWTB fibers > GF.

Figure 11e also shows improved ratios of freeze–thaw splitting tension strength for each fiber-modified binder. PF had the highest improvement (13.5%) among base asphalts and was followed by BF (7.5%), RWTB fibers (8.1%), and GF (5.9%). Improvement was mostly lower for SBS-modified asphalts with PF > BF > RWTB fibers > GF.

As can be observed from market survey [40], reclaimed RWTB fibers can be purchased at about 1 yuan/m³, and conventional AC-13 and SBS-modified AC-13 mixes can be bought at about 0.4 yuan/m³ and 0.5 yuan/m³, respectively. GF, PF, and BF are more costly with prices 5.2 yuan/m³, 7.8 yuan/m³, and 12.8 yuan/m³, respectively. Considering optimal fiber content and mixture density, engineering costs per cubic meter for fiber-reinforced asphalt mixtures were calculated and given in Figure 11f.

4.2. Analytic Hierarchy Process (AHP)

Step one: Decompose the decision problem into multiple levels, typically including the goal layer G, the criterion layer C = {C1,C2,C3,…,Cn} and the alternative layer A = {A1,A2,A3,…,An};

Step two: Construct a judgment matrix. For each criterion, a pairwise comparison judgment matrix A = [aij] is constructed, as shown in Equation (2), where aij represents the importance of criterion i relative to criterion j. The diagonal elements of the matrix are all 1 [41].

$$A=\left[\begin{array}{ccc}1& \dots & a_{1n}\\ ⋮& \ddots & ⋮\\ {\displaystyle \frac{1}{a_{1n}}}& \dots & 1\end{array}\right]$$

(2)

Step three: Calculate the eigenvector of the judgment matrix A and normalize it to obtain the weight vector W. Use Equation (3) to calculate the sum of each column, then use Equation (4) to normalize the judgment matrix, and finally calculate the mean value of each normalized row to obtain the weight of each criterion.

$$s_j={\sum }_{i=1}^n{a}_{ij}$$

(3)

$$\widetilde{a_{ij}}={\displaystyle \frac{a_{ij}}{s_j}}$$

(4)

$$w_i={\displaystyle \frac{1}{n}}{\sum }_{i=1}^n\widetilde{a_{ij}}$$

(5)

Step four: Perform a consistency check on the obtained weights. Here, RI is the random consistency index, determined based on the scale n of the judgment matrix. If CR < 0.1, the judgment matrix is considered consistent, and the obtained weights can be used for this evaluation.

$$CI={\displaystyle \frac{{\lambda }_{max}-n}{n-1}}$$

(6)

$$CR={\displaystyle \frac{CI}{RI}}$$

(7)

4.3. Analysis of Results

Based on the comparative analysis of pavement performance, BF exhibited the best high-temperature stability, while GF performed the worst. In terms of cost, GF was the most expensive, whereas PF had the lowest cost. To comprehensively compare the performance of the 10 fiber-reinforced asphalt mixtures, the Analytic Hierarchy Process (AHP) was introduced. This method enables an effective comprehensive evaluation of the performance of different fibers incorporated into both base and modified asphalt mixtures [42]. The results are shown in

Table 13. Judgment matrix and corresponding weights for each criterion.

Pairwise Comparison Matrix	Cost	Bending Tensile Strain	Residual Stability	Dynamic Stability	Row Sum Product	Eigenvector	Normalization
Cost	1.00	0.33	0.50	0.33	0.06	0.49	0.11
Bending tensile strain	3.00	1.00	2.00	1.00	6.00	1.57	0.35
Residual stability	2.00	0.50	1.00	0.50	0.50	0.84	0.19
Dynamic stability	3.00	1.00	2.00	1.00	6.00	1.57	0.35

,

Table 14. Consistency checklist.

Criteria Layer	Normalization	AW	AW/W	Maximum Eigenvalue	CI	RI	CR
Cost	0.11	0.44	4.02	4.01	0.00	0.90	0.004
Bending tensile strain	0.35	1.41	4.01
Residual stability	0.19	0.76	4.02
Dynamic stability	0.35	1.407	4.01

and

Table 15. Comprehensive Scoring Results and Ranking.

Asphalt Mixture Types	Cost	Bending Tensile Strain	Residual Stability	Dynamic Stability	Overall Evaluation Score	Ranking
SBS/BF AC-13	1.71	1.19	0.99	1.12	1.19	1
SBS/RWTB AC-13	0.66	1.03	0.31	0.95	0.82	2
SBS/GF AC-13	0.79	0.86	0.19	0.91	0.75	3
SBS/PF AC-13	0.60	0.20	0.82	1.10	0.68	4
SBS AC-13	0.40	0.36	−0.17	0.62	0.36	5
BF AC-13	0.05	0.45	0.82	−0.82	0.03	6
RWTB AC-13	−1.00	−0.46	−0.27	−0.96	−0.66	7
GF AC-13	−0.87	−0.55	−0.50	−0.97	−0.72	8
PF AC-13	−1.07	−1.79	0.26	−0.93	−1.02	9
AC-13	−1.26	−1.29	−2.47	−1.01	−1.41	10

, where the CR value in Table 14 is 0.004, far below 0.1, indicating that the normalized data in Table 15 is valid and can be further analyzed. According to the results calculated using the AHP method: For SBS-modified asphalt mixtures, the ranking of the overall improvement is BF > RWTB > GF > PF. For base asphalt mixtures, the ranking is also BF > RWTB > GF > PF. Both base and modified asphalt mixtures with fiber reinforcement demonstrated better technical performance than mixtures without fibers. SBS-modified asphalt mixtures ranked higher due to their cost-effectiveness and simultaneous improvement in pavement performance.

Although the overall performance of RWTB fibers ranked second, slightly weaker than SBS/BF AC-13, as a recycled waste material, the use of RWTB fiber-reinforced mixtures contributes to addressing environmental pollution caused by decommissioned wind turbine blades. This aligns positively with China’s green and low-carbon development policies.

5. Conclusions

(1)

RWTB fibers are cylindrical glass fiber bundles with diameters of approximately 10–15 μm and lengths exceeding 200 μm. These fibers exhibit uniform thickness and an interwoven structure, with residual resin microparticles adhered to their surfaces. No new chemical reactions or functional groups were identified between the fibers and the asphalt matrix. The reinforcing and bridging effects of the fibers promote internal network formation, thereby enhancing the toughness, elasticity, and overall mechanical performance of asphalt mixtures.

(2)

The optimal preparation parameters for RWTB fiber-reinforced asphalt mixtures were identified as follows: fiber preheating at 120 °C, aggregate heating at 170 °C, mixing temperature at 180 °C, and a total mixing duration of 180 s (60 s for each stage: dry mixing with aggregate, wet mixing after asphalt addition, and final mixing after mineral filler incorporation). The optimal fiber particle size range was determined to be 0.3–1.18 mm.

(3)

Based on gray relational analysis and the entropy weighting method, the optimal RWTB fiber content was determined to be 0.15 wt%. Under this dosage, the base asphalt mixtures showed performance improvements of 27.1% in dynamic stability, 23.8% in low-temperature flexural tensile strain, 9.9% in Marshall stability after immersion, and 8.1% in freeze–thaw splitting strength. Corresponding enhancements in SBS-modified mixtures were 18.3%, 12.9%, 2.3%, and 2.5%, respectively.

(4)

A comprehensive performance evaluation using the Analytic Hierarchy Process (AHP) revealed the following ranking of fiber-reinforced asphalt mixtures: BF > RWTB > GF > PF. RWTB fibers outperformed both GF and PF, ranking second only to BF, indicating their high potential for enhancing asphalt pavement performance.

(5)

This study presents a systematic and practical approach to the high-value utilization of recycled wind turbine blade (RWTB) materials in road engineering. It offers a feasible technical solution for the resource recovery of wind power composites, contributing to the reduction, harmless disposal, and reuse of retired turbine blades. These findings carry significant environmental and engineering implications and support the development of green, low-carbon, and sustainable transportation infrastructure aligned with global sustainability goals.

Despite these promising results, certain limitations remain. Future research should aim to:

(1)

investigate the mechanisms by which RWTB fibers enhance low-temperature cracking resistance, including potential surface treatments to improve adhesion;

(2)

conduct long-term durability tests—such as aging, fatigue, and moisture resistance tests—to evaluate performance throughout the pavement lifecycle;

(3)

implement pilot-scale or field applications to validate the laboratory results under real-world operating conditions.