Effects of preprocessing methods on the road performance of high reclaimed asphalt pavement recycled hot-mix asphalt

ABSTRACT

High reclaimed asphalt pavement (RAP) contents in recycled hot-mix asphalt (RHMA) provide economic and environmental benefits but face performance and compaction challenges. This study compared refined decomposition (RD) and roll crushing (RC) for 0%, 30%, 50%, and 70% RAP mixtures. Mix design, gyratory-compaction indices, and performance tests were conducted. Results show that, Compared with RC-RAP, RD-RAP mixtures showed higher Marshall stability, lower air voids, and higher voids filled with asphalt (VFA). At equal RAP contents, optimum asphalt content (OAC) decreased by 0.1%–0.3%, and the compaction energy index (CEI) fell by 13.95%–31.58%, with greater reduction at higher RAP. In terms of road performance, RD-RAP exhibited slightly lower rutting resistance but higher moisture and cracking resistance: TSR increased by 6.37%–10.41%, fracture energy by 11.59%–18.14%. RD-RAP at 70% outperformed RC-RAP at 50%. The dynamic modulus of RD-RAP increased monotonically with RAP, while RC-RAP peaked at 50%. Overall, refined decomposition alleviates compaction and cracking limitations, enabling RAP content up to 70% and offering technical guidance for high-RAP application.

Keywords:
RAP; Preprocessing methods; Refined decomposition; Performance test

1. INTRODUCTION

In China, highway construction is gradually shifting from large-scale new construction to maintenance and rehabilitation, which generates a substantial amount of RAP that requires recycling [¹,²,³]. Statistics indicate that approximately 790 million tons of RAP are produced annually, and this figure is expected to increase significantly in the coming years [⁴]. Given the considerable economic and environmental benefits of effective RAP recycling, improving the utilization rate of RAP has become a major research focus in recent years [⁵,⁶,⁷]. Previous studies have shown that when RAP content is within 15%–20%, RHAM exhibits performance comparable to conventional asphalt mixtures. However, when RAP content is further increased, it often exerts detrimental effects on the performance of RHAM [⁸, ⁹].

The limitations of RAP utilization primarily stem from two aspects: the insufficient viscoelasticity of aged binders and the high variability of RAP [¹⁰]. Asphalt pavements undergo inevitable aging during long-term service, as the binder is continuously exposed to environmental conditions. The aging process leads to an increase in binder viscosity and stiffness, which may enhance the high-temperature performance of the mixture; however, the accompanying rise in brittleness results in a deterioration of low-temperature cracking resistance. In addition, the accumulation of polar functional groups in the aged binder weakens its adhesion to aggregates, thereby increasing the risk of moisture-induced debonding. Under repeated traffic loading and environmental stresses, microcracks may also develop within the aged binder, further reducing the fatigue durability of the mixture [¹¹,¹²,¹³]. Moreover, variability in RAP is another critical constraint. This variability originates from factors such as pavement service age, degree of damage, and the heterogeneous blending of RAP obtained from different pavement layers [¹⁴]. Excessive RAP incorporation may lead to discrepancies between the actual and designed gradation, thereby resulting in unstable performance and unreliable quality of RHAM [¹⁵].

Compaction remains one of the most critical challenges in the practical application of high-RAP hot-mix recycled asphalt mixtures. The presence of aged binder with elevated viscosity and stiffness significantly reduces mixture workability, leading to higher compaction effort and difficulty in achieving target air-void levels during construction. Previous studies have shown that inadequate blending between aged and virgin binders, together with RAP agglomeration, further exacerbates compactability issues by creating locally stiff zones within the mixture [¹⁶, ¹⁷], which may increase the likelihood of quality-related problems during field construction.

To address these issues, researchers have turned their attention to RAP preprocessing technologies, aiming to enhance RAP utilization efficiency [¹⁸, ¹⁹]. Common preprocessing techniques can be classified into three categories: chemical, physical, and biological separation. Chemical separation dissolves aged binder using solvents to achieve efficient separation of asphalt from aggregates; however, solvents are costly and pose potential risks to human health and the environment [²⁰]. Physical separation, typically involving roll crushers, jaw crushers, or impact crushers, is less costly but may damage aggregate strength and suffers from relatively low efficiency. Biological separation, which relies on microbial or enzymatic activity to degrade and strip binders from aggregates, remains at an early stage of development and has limited efficiency [¹⁹]. In addition to these methods, refined decomposition technology, originally derived from oil–sand separation in the petroleum industry, has recently been applied to asphalt pavement recycling and demonstrated promising potential in controlling RAP gradation variability [²¹].

Refined decomposition technology shows promising application prospects in the field of asphalt pavement engineering. Ai et al. [⁹] prepared RHAM with varying RAP contents using refined decomposition (RD-RAP), analyzed the corresponding performance, and developed a predictive model for dynamic modulus. Yu et al. [²²] assessed the separation efficiency of refined decomposition and its effects on aggregate properties. Tang et al. [²³] employed digital image processing (DIP) to evaluate the influence of RD-RAP on aggregate uniformity. Wang et al. [²⁴] investigated the effects of frequency parameters during refined decomposition on RAP gradation, asphalt content, and aggregate properties, and further determined the maximum feasible RAP content. Qu et al. [¹⁸] optimized refined decomposition parameters using orthogonal design and examined the fatigue performance of RHAM produced with RD-RAP. Collectively, these studies demonstrated the significant potential of refined decomposition in enabling higher RAP utilization.

Building upon these findings, this study prepared RHAM mixtures with RAP contents of 0%, 30%, 50%, and 70%, using both RD and RC as preprocessing methods. A comprehensive experimental program, including rutting, freeze–thaw splitting, SCB, dynamic modulus, and gyratory compaction tests, was conducted to compare the effects of the two preprocessing approaches on the volumetric properties, road performance, and compaction characteristics of RHAM. The results are expected to provide insights for the development of effective preprocessing technologies and the performance evaluation of recycled asphalt mixtures.

2. RAP PREPROCESSING METHODS

According to the specifications for Technical Specifications for Highway Asphalt Pavement Recycling (JTG/T 5521—2019) [²⁵], RAP must be pretreated through screening and crushing in plant hot recycling to reduce agglomeration. In this study, two preprocessing methods, namely RD and RC, were employed. The detailed procedures of the two methods are described as follows.

2.1. Refined decomposition

The core process of RD is based on high-frequency rotary centrifugation, in which the aged asphalt mortar on RAP undergoes glassy brittleness under vibration. During the centrifugal impact, RAP particles collide with the inner wall of the centrifuge, causing substantial detachment of aged mortar from the coarse aggregates, thereby achieving separation between the aged binder and RAP aggregates. The detached aged mortar is transferred into the fine aggregate fraction, resulting in RAP coarse aggregates with minimal aged asphalt and fine aggregates enriched with binder [⁶]. The detailed procedure is as follows: untreated RAP is first conveyed to a screening system. Oversized RAP particles larger than 22 mm are crushed and then reintroduced into the screening process. RAP fractions of 5–22 mm are fed into a rotary centrifuge for decomposition, where they are separated into three gradations: 0–5 mm (after refined decomposition), 5–11 mm, and 11–22 mm. Meanwhile, the 0–5 mm fraction obtained directly from screening is transported via conveyor to the finished stockpile. The flowchart of RD method is shown in Figure 1. Technical specifications of the RD equipment is shown in Table 1.

Figure 1
Flowchart of RD method.

Table 1
Technical specifications of the RD equipment.

MODEL	RAPR120
Total machine power	370 Kw
Rated production capacity	120 t/h
Maximum feed size	40 mm
Discharge particle size	≤25 mm

2.2. Roll crushing

RC involves crushing oversized RAP particles with a roll crusher, followed by gradation using a vibrating screen. The specific procedure is as follows: untreated RAP is first conveyed to the vibrating screen and separated into three fractions: 0–8 mm, 8–12 mm, and 12–22 mm. RAP particles larger than 22 mm are subjected to roll crushing and subsequently reintroduced into the screening system [²⁶]. The flowchart of RC method is shown in Figure 2. Technical specifications of the RC equipment is shown in Table 2.

Figure 2
Flowchart of RC method.

Table 2
Technical specifications of the RC equipment.

MODEL	RPS100
Total machine power	121 Kw
Rated production capacity	100 t/h
Maximum feed size	800 mm
Discharge particle size	≤32 mm

3. MATERIALS

3.1. RAP

The RAP used in this study was collected from the Wensha section of the G70 Fuyin Expressway and pretreated at the Nancheng base of Jiangxi Communications Engineering Group using RD and RC. The RD-RAP was classified into three fractions: A (0–5 mm), B (5–11 mm), and C (11–22 mm). The RC-RAP was divided into A (0–8 mm), B (8–12 mm), and C (12–22 mm). According to the Specifications for Test Methods of Asphalt and Asphalt Mixtures for Highway Engineering (JTG E20-2011), each RAP sample was subjected to an ignition test. After combustion, the asphalt content of RAP was determined, and the recovered aggregates were washed and sieved to measure the RAP aggregate gradation. The aggregate gradations and binder content of the two types of RAP are presented in Tables 3 and 4, while the fundamental properties of the aged binder are provided in Table 5.

Table 3
Aggregate gradation of RD-RAP.

SIEVE SIZE (mm)	PASSING PERCENTAGE of RAP MATERIALS (%)
	0–5 mm	5–11 mm	11–22 mm
26.5	100.00	100.00	100.00
19	100.00	100.00	99.50
16	100.00	100.00	98.89
13.2	100.00	100.00	94.99
9.5	100.00	98.97	45.07
4.75	96.35	32.72	12.39
2.36	78.82	12.41	10.99
1.18	62.52	10.60	9.60
0.6	45.78	8.67	7.87
0.3	33.47	6.90	6.33
0.15	25.05	5.57	5.15
0.075	19.94	4.44	4.17
Binder content (%)	6.41	2.39	2.19

Table 4
Aggregate gradation of RC-RAP.

SIEVE SIZE (mm)	PASSING PERCENTAGE of RAP MATERIALS (%)
	0–8 mm	8–12 mm	12–22 mm
26.5	100.00	100.00	100.00
19	100.00	100.00	97.08
16	100.00	100.00	88.20
13.2	100.00	99.80	74.59
9.5	99.90	95.75	42.67
4.75	87.79	31.47	26.43
2.36	69.80	20.40	20.61
1.18	55.23	16.59	16.63
0.6	39.61	13.03	12.70
0.3	27.94	10.03	9.65
0.15	20.60	7.67	7.32
0.075	16.12	6.06	5.87
Binder content (%)	6.04	4.08	3.22

Table 5
Fundamental properties of aged binder.

TEST ITEMS	TEST VALUE	METHODS
(25 °C, 100 g, 5 s) Penetration (0.1 mm)	31.5	T0604
(5 cm·min-1, 10 °C) Ductility (cm)	1.7	T0605
Softening point (°C)	74.6	T0606
Viscosity at 135 °C (Pa·s)	1.72	T0625

3.2. Original aggregate and filling

Limestone was used as the virgin aggregate, and limestone powder was adopted as the filler. Their physical properties were determined in accordance with the Specifications for Test Methods of Aggregate for Highway Engineering (JTG E42-2005) [²⁷], and the corresponding results are summarized in Tables 6 and 7.

Table 6
Physical characteristics of limestone aggregates.

SIEVE SIZE (mm)	TEST ITEMS
	BULK RELATIVE DENSITY	APPARENT RELATIVE DENSITY	WATER ABSORPTION RATE (%)
19	2.691	2.716	0.41
16	2.702	2.720
13.2	2.691	2.716	0.54
9.5	2.690	2.718
4.75	2.685	2.178	0.59
2.36	2.646	2.719
1.18	–	2.687	0.73
0.6	–	2.657
0.3	–	2.663
0.15	–	2.635
0.075	–	2.618
Requirements	≥2.6	–	≤3

Table 7
Filling technical index.

TEST ITEMS		TEST VALUE	REQUIREMENT	STANDARD
Water absorption rate (%)		0.45	≤1	–
Relative density		2.679	≥2.5	T0352
Appearance		No clumping	No clumping	–
Hydrophilicity		0.62	<1	T0353
Passing percentage (%)	<0.6 mm	100	100	T0353
	<0.15 mm	93.9	90–100
	<0.075 mm	93.4	75–100

3.3. Virgin binder

The virgin binder used in this study was SBS-modified asphalt supplied by a materials company in Jiangxi Province. Its fundamental properties were tested in accordance with the specifications for Standard Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering (JTG E20-2011) [²⁸], and the results are presented in Table 8.

Table 8
Physical characteristics of limestone aggregates.

TEST ITEMS	TEST VALUE	REQUIRMENT	METHODS
(25 °C, 100 g, 5 s) Penetration (0.1 mm)	72.1	60–80	T0604
Softening point (°C)	64.5	≥55	T0606
(5 cm·min-1, 10 °C) Ductility (cm)	46	≥30	T0605
(135 °C) Viscosity (Pa·s)	1.56	≤3	T0626
After RTFOT 85 min
Change in quality (%)	0.4	≤ ±1	T0610
(25 °C) Penetration ratio (%)	74.6	≥60	T0610
(5 cm·min-1, 10 °C) Ductility (cm)	23.9	≥20	T0610

3.4. Rejuvenator

The rejuvenator used in this study was Evoflx8182, produced by Ingevity Corporation, and its properties are summarized in Table 9. Previous studies have suggested that the optimum dosage of rejuvenators typically ranges from 4% to 6% of the aged binder [⁴, ²⁹, ³⁰]. Considering that this work primarily focuses on the influence of two preprocessing methods on the road performance of recycled asphalt mixtures, and in order to minimize additional variables, the rejuvenator content was uniformly selected as 5% of the aged binder based on the properties of both the binder and the rejuvenator.

Table 9
Basic properties of SBS-modified asphalt.

TEST ITEMS	TEST VALUE	METHODS
(60 °C) Viscosity (cst)	9532	T0619
Flash point (°C)	240.4	T0633
Saturates (%)	21.5	T0618
Aromatics (%)	41.5	T0618

4. EXPERIMENTS AND METHODS

4.1. Mix proportion design of recycled asphalt mixture

Based on the gradation range of AC-20 recommended by the specification for Technical Specification for Construction of Highway Asphalt Pavements (JTG E40-2004) [³¹], a target gradation was determined to investigate the effects of the two preprocessing methods on the overall performance of recycled asphalt mixtures. To minimize the influence of gradation discrepancies, seven groups of mixtures (0% RAP, 30% RD-RAP, 50% RD-RAP, 70% RD-RAP, 30% RC-RAP, 50% RC-RAP, and 70% RC-RAP) were prepared using the same target gradation, with virgin aggregates adjusted in single sizes to match this target. The upper and lower limits, median, and the target gradation are illustrated in Figure 3.

Figure 3
Gradation curve of AC-20 recycled asphalt mixture.

The OAC of each recycled asphalt mixture was determined using the Marshall design method. According to the requirements of JTG/T 5521-2019, the heating and mixing temperatures were set as follows: RAP was heated at 120 °C for 2 h, SBS-modified binder at 165 °C for 2 h, and virgin aggregates and filler at 180 °C for 4 h. During mixing, the temperature of the mixing pan was maintained at approximately 165 °C. RAP and the rejuvenator were first added and mixed for 60 s, followed by virgin aggregates for another 60 s, after which SBS-modified binder and limestone filler were introduced sequentially. The prepared mixtures were then compacted at 155 °C using a Marshall hammer. In accordance with JTG E40-2004, the volumetric parameters, stability, and flow values of the specimens were measured. The air voids, voids in mineral aggregate (VMA), and VFA were subsequently calculated to determine the OAC for each mixture. The specimen preparation procedure is illustrated in Figure 4.

Figure 4
Specimen preparation procedure.

4.2. Gyratory compaction test

To evaluate the compaction characteristics of each recycled asphalt mixture, gyratory compaction tests were carried out in accordance with the Chinese specification JTG E20-2011, with four specimens prepared for each mixture. The compaction performance was assessed using the Compaction Energy Index (CEI) and the Transportation Densification Index (TDI), as shown in Figure 5. CEI represents the compaction energy required from the paving state to the target density. In this study, CEI was defined as the total work performed from the 8th gyration to the point where the degree of compaction reached 96%. TDI simulates the secondary densification process of asphalt mixtures under traffic loading after the pavement is opened to service and was defined as the total work performed when the degree of compaction increased from 96% to 98%. A lower CEI indicates that the asphalt mixture is more compactable, suggesting favorable compaction characteristics. Conversely, a higher TDI implies greater resistance to permanent deformation after opening to traffic. The calculation formulas for these two indices are given in Equation (1) and Equation (2).

Figure 5
Gyratory compaction test: (a) gyratory compactor; (b) compaction curve.

(1) $$CEI={\displaystyle \sum _{n=8}^{n_i=N_{G=96\%}}(G_{ni}}-G_8)d(n_i)$$

(2) $$TDI={\displaystyle \sum _{n_i=N_{G=96\%}}^{n_i=N_{G=98\%}}(}{G}_{ni}-96)d(n_i)$$

Where: N_G=96% is the number of gyrations corresponding to a degree of compaction of 96%; G_ni is the compaction curve function; G₈ is the degree of compaction at the 8th gyration; N_G=98% is the number of gyrations corresponding to a degree of compaction of 98%.

4.3. High-temperature rutting test

The rutting test is commonly employed to evaluate the high-temperature stability of recycled asphalt mixtures, with dynamic stability (DS) used as the indicator of rutting resistance. In accordance with JTG E40-2004, rutting slabs with dimensions of 300 mm × 300 mm × 50 mm were fabricated using a rutting slab compactor, and four slabs were prepared for each mixture. After cooling for 24 h, the specimens were tested using a wheel-tracking device at a temperature of 60 °C, with a wheel load of 0.7 MPa and a loading duration of 1 h. For each type of recycled mixture, four parallel tests were conducted, and the average value was taken as the result, as shown in Figure 6. A higher DS value indicates superior rutting resistance, and its calculation is given in Equation (3).

Figure 6
High-temperature rutting test: (a) rutting slab compactor; (b) wheel-tracking device.

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

Where:t₁ is the 45th minute from the start of the test; t₂ is the 60th minute from the start of the test; d₁ is the deformation at t₁ (mm); d₁ is the deformation at t₂ (mm); C₁ and C₂ are test coefficients, both taken as 1; and N is the loading frequency of the wheel passes, 42 cycles/min.

4.4. Freeze–thaw splitting test

The freeze–thaw splitting test is commonly used to evaluate the moisture stability of recycled asphalt mixtures, with the tensile strength ratio (TSR) serving as a quantitative indicator of resistance to moisture-induced damage. In accordance with JTG E40-2004, Marshall specimens with a diameter of 101.6 mm and a height of 63.5 mm were prepared. The control specimens were stored at room temperature, while the conditioned specimens were vacuum-saturated, frozen at –16 °C for 16 h, and subsequently immersed in a 60 °C water bath for 24 h. Both groups were then conditioned in a 25 °C water bath for 2 h prior to the splitting test to determine tensile strength. The experimental procedure is illustrated in Figure 7 The TSR was calculated using Equations (2)–(4).

Figure 7
Freeze–thaw splitting test: (a) freezing at –16 °C for 16 h; (b) water bath at 60 °C for 24 h.

(4) $$S_{T1}=\frac{0.006287P_{T1}}{h_1}$$

(5) $$S_{T2}=\frac{0.006287P_{T2}}{h_2}$$

(6) $$ITSR=\frac{\overline{S_{T1}}}{\overline{S_{T2}}}\times 100\%$$

Where: S_T1 is the splitting tensile strength of the control specimens without freeze–thaw conditioning (MPa); S_T2 is the splitting tensile strength of the conditioned specimens after freeze–thaw cycles (MPa); P_T1 is the maximum load of a single specimen during the test (N); h is the specimen height (mm);$\overline{S_{T1}}$ is the average splitting tensile strength of the control group (MPa); $\overline{S_{T2}}$ is the average splitting tensile strength of the conditioned group (MPa); and TSR is the tensile strength ratio (%).

4.5. SCB test

The SCB test is commonly used to evaluate the low-temperature cracking resistance of recycled asphalt mixtures, with fracture energy and fracture toughness adopted as the primary indicators. In accordance with JTG E40-2004, cylindrical specimens with dimensions of 150 mm × 140 mm were first fabricated using a Superpave gyratory compactor. After curing at room temperature for 24 h, the specimens were cut with a large saw into semi-circular samples with a diameter of 150 mm, a radius of 75 mm, and a thickness of 50 mm, containing a pre-cut notch of 15 mm in depth and 1.5 mm in width. For each recycled mixture, four parallel tests were conducted, yielding a total of seven groups. Finally, the SCB tests were carried out using a multifunctional asphalt mixture testing machine at –10 °C with a loading rate of 1.5 mm/min. res. The experimental procedure is illustrated in Figure 8. The G_f and K_IC were calculated according to Equations (7)–(9).

Figure 8
SCB test: (a) semi-circular specimen; (b) loading process.

(7) $$W_f={\displaystyle \int Pdu}$$

(8) $$G_f=\frac{W_f}{Are{a}_{lig}}$$

(9) $$K_{IC}=\frac{P}{2rt}\times Y_I\times \sqrt{\pi a}$$

Where: W_f is the fracture work (J); P is the applied load (N); uis the load-line displacement (mm); G_f is the fracture energy (J·m⁻²); Area_lig is the ligament area (mm²); Area_lig= (r-a)·t; r is the specimen radius (mm) ; a is the notch depth (mm), t is the specimen thickness (mm); K_IC is the fracture toughness (MPa·m^-2); Y_I is the standard stress intensity factor.

4.6. Test dynamic modulus test

The dynamic modulus is a key parameter reflecting the viscoelastic behavior of recycled asphalt mixtures and plays an important role in asphalt pavement design. According to the specification, cylindrical specimens with a diameter of 150 mm and a height of 170 mm were first fabricated using a Superpave gyratory compactor. After cooling at room temperature for 24 h, core samples with a diameter of 100 mm and a height of 150 mm were obtained for testing. For each mixture, three parallel tests were conducted. The test temperature was set at 20 °C, the applied strain was 100 × 10⁻⁶, and the loading frequencies were 0.1, 0.5, 1, 5, 10, and 25 Hz. The experimental procedure is illustrated in Figure 9. The dynamic modulus was calculated according to Equation (10).

Figure 9
Dynamic modulus test.

(10) $$\left|E^{*}\right|=\frac{{\sigma }_0}{{\epsilon }_0}$$

Where: |E*| is the dynamic modulus (MPa); σ₀ is the axial stress (MPa); ε₀ is the measured strain (µε).

5. RESULTS AND DISCUSSION

5.1. Results of mix proportion design of recycled asphalt mixture

In the mix design, five sets of Marshall specimens with asphalt contents spaced at 0.3% intervals were prepared for each RHMA to determine the OAC. Taking the mixture with 70% RAP as an example, the design results are illustrated in Figure 10, while the OAC values of all mixtures are summarized in Table 10. As shown in Figure 10, at the high RAP content of 70%, The bulk specific gravity of both mixtures first increased and then de-creased with binder content, while RD-RAP consistently achieved higher values than RC-RAP, indicating improved compactness and structural stability. Marshall stability exhibited a similar trend, but the peak stability of RD-RAP reached 17.89 KN, which was higher than that of RC-RAP (14.86 KN), demonstrating the superior load-bearing capacity. The air voids decreased with increasing binder content for both groups, but RD-RAP showed lower void ratios, reflecting denser internal structures. In terms of VFA, RD-RAP presented consistently higher values than RC-RAP, which suggests more effective binder filling.

Figure 10
Marshall test results of 70% RAP mixture: (a) bulk density; (b) stability; (c) air voids; (d) flow value; (e) VFA; (f) VMA.

Table 10
Optimum asphalt contents of different RHMA.

ASPHALT CONTENT	TYPES OF RHMA
	0%	30% (RD)	30% (RC)	50% (RD)	50% (RC)	70% (RD)	70% (RC)
OAC (%)	3.8	3.9	4.0	4.2	4.4	4.3	4.6
Virgin binder content (%)	–	2.5	2.8	2.3	2.6	2.0	2.2

Table 10 further shows that, regardless of the preprocessing method, the OAC in-creased with increasing RAP content. This trend can be attributed to the fact that not all aged asphalt participates in blending with virgin asphalt as the RAP content rises. The presence of non-blend able “black rock” reduces the overall degree of asphalt blending, thus requiring additional virgin asphalt to meet design requirements [³²]. Moreover, under the same RAP content, RD-RAP mixtures consistently required less OAC than RC-RAP mixtures. Specifically, for 30%, 50%, and 70% RAP, the OACs of RD-RAP mixtures were 3.9%, 4.2%, and 4.3%, respectively, while those of RC-RAP mixtures were 4.0%, 4.4%, and 4.6%, with differences ranging from 0.1% to 0.3%. This phenomenon can be explained by the refined decomposition process, which reduces RAP agglomeration, promotes more effective blending between aged and virgin asphalt, and facilitates better diffusion of the rejuvenator into aged asphalt, thereby enhancing performance recovery and reducing the demand for additional virgin asphalt. This mechanism also explains how the preprocessing method influences the volumetric indices. Aged binder typically exhibits higher viscosity, and more effective restoration results in a more uniform and sufficient reduction in its viscosity. Through the elimination of RAP agglomeration, the RD method allows the aged binder to be more adequately exposed and more easily restored toward its original viscosity. This improvement enhances the mixture’s ability to achieve the target compaction density. Consequently, under the Marshall mix design framework, this manifests as a lower optimal asphalt content and corresponding variations in the volumetric parameters.

5.2. Results of gyratory compaction test

The compaction test results of RHMA with different RAP contents and preprocessing methods are shown in Figure 11. For CEI, values increased with RAP content. In RD-RAP mixtures, CEI rose from 146.7 at 0% RAP to 148.7, 183.3, and 272.2 at 30%, 50%, and 70% RAP, corresponding to growth rates of 0.095%, 24.94%, and 85.55%, respectively. Com-pared with RC-RAP mixtures, CEI at the same RAP level was reduced by 13.95%–31.58%. These results indicate that the compactibility of RHMA is strongly affected by RAP content. This can be attributed to the aged asphalt in high RAP contents, which increases mixture stiffness, weakens binder–aggregate adhesion, and raises particle friction, thereby requiring greater compaction energy [⁵, ³³]. In contrast, RD-RAP mixtures not only required lower OAC but also exhibited better compactibility at higher RAP contents. Specifically, the compactibility of RD-70% RAP mixtures was comparable to that of RC-50% RAP mixtures. Overall, refined decomposition proved more effective in alleviating the compaction difficulties typically associated with high RAP contents.

Figure 11
Results of gyratory compaction test.

For TDI, a similar trend was observed, with higher RAP contents resulting in larger TDI values. This outcome is reasonable because TDI characterizes the resistance to post-construction densification under traffic loads, and as confirmed by the rutting test results, the incorporation of RAP increases the stiffness and deformation resistance of RHMA. Consequently, more compaction energy is required for the degree of compaction to rise from 96% to 98%, leading to higher TDI values with increasing RAP content. However, excessive RAP contents also reduce mixture ductility, thereby raising the risk of brittle fracture.

5.3. Results of high-temperature rutting test

Figure 12 presents the rutting test results of RHMA with different RAP contents under the two preprocessing methods. The red dashed line represents the minimum requirement specified in the standard, namely 2800 passes/mm, and all mixtures satisfied this requirement. As shown in Figure 11, regardless of the preprocessing method, the DS of RHMA increased with higher RAP content. This may be attributed to the greater proportion of aged binder at higher RAP levels, since aged binder possesses higher stiffness compared with virgin binder, thereby enhancing the rutting resistance of RHMA. It can also be observed from Figure 11 that, at the same RAP content, RHMA prepared with RC-RAP exhibited higher DS values than those prepared with RD-RAP, with the difference becoming more pronounced at 70% RAP. Specifically, for RHMA prepared with RD-RAP, the DS increased from 4957 passes/mm at 0% RAP to 5478 passes/mm at 30%, 6347 passes/mm at 50%, and 6847 passes/mm at 70%, corresponding to increases of 10.51%, 28.04%, and 38.13%, respectively. In contrast, the DS of RHMA prepared with RC-RAP increased by 12.75%, 29.70%, and 56.91% at RAP contents of 30%, 50%, and 70%, respectively. The difference between the two groups may be explained by the fact that RD reduces the agglomeration of RAP, thereby allowing more effective blending between the rejuvenator and the aged binder, which facilitates better restoration of the aged binder properties. As a result, RHMA prepared with RD-RAP exhibits relatively lower DS values than that prepared with RC-RAP at the same RAP content [³⁴].

Figure 12
High-temperature rutting test results.

5.4. Results of freeze–thaw splitting test

Figure 13 shows the freeze–thaw splitting results of RHMA with different RAP con-tents and preprocessing methods, where the red dashed line indicates the minimum requirement specified by the standard. After one freeze–thaw cycle, the TSR of both RC-RAP and RD-RAP mixtures decreased with increasing RAP content. When the RAP content in RC-RAP mixtures reached 50%, their water stability no longer satisfied the specification, whereas RD-RAP mixtures still met the requirement even at 70% RAP. The decline in water stability with RAP addition can be attributed to the limited rejuvenating effect of the additive, which cannot fully restore the aged asphalt or achieve complete blending with virgin asphalt, thus leaving weak interfaces [³⁵]. During freeze–thaw cycles, water penetrates these weak zones, and the expansion upon freezing aggravates the interfacial damage, further reducing RHMA durability. Moreover, under identical RAP contents, RD-RAP mixtures exhibited 6.37%–10.41% higher TSR values than RC-RAP mixtures. This improvement is explained by the refined decomposition process, which facilitates the detachment of aged asphalt from RAP particles, reduces agglomeration, and consequently minimizes weak interfaces [²², ³⁵, ³⁶]. To ensure RHMA quality, it is recommended that RC-RAP contents do not exceed 50%, while RD-RAP mixtures can still maintain relatively high re-liability even at 70% RAP.

Figure 13
Results of freeze–thaw splitting test.

5.5. Results of semi-circular bending test

Figure 14 presents the SCB test results of RHMA with different RAP contents under the two preprocessing methods. At –10 °C, the Fracture energy are shown in Figure 14(a). For both RC-RAP and RD-RAP mixtures, Fracture energy decreased with increasing RAP content, indicating that the incorporation of RAP reduced the low-temperature cracking resistance of RHMA. Compared with RAP-free mixtures, the Fracture energy of RD-RAP mixtures decreased from 1316.9 J/m² at 0% RAP to 1034.1 J/m² at 30%, 859.4 J/m² at 50%, and 765.9 J/m² at 70%, corresponding to reductions of 21.47%, 34.74%, and 41.84%, respectively. For RC-RAP mixtures, the reductions were 29.63%, 44.54%, and 51.35% at RAP levels of 30%, 50%, and 70%, respectively. This phenomenon can be explained by the fact that at –10 °C, RHMA exhibits a high-modulus and brittle state, which limits its deformation capacity [⁵]. In addition, the agglomeration of RAP creates internal voids in RHMA. Under low-temperature loading, these voids act as stress concentration zones, leading to crack propagation along the weak regions and consequently reducing fracture resistance. It is further observed from Figure 14(a) that, at the same RAP content, the fracture energy of RD-RAP mixtures increased by 11.59–18.14% compared with that of RC-RAP mixtures. This difference can be attributed to the reduction in RAP agglomeration and variability achieved by RD [⁹, ³⁵], which enhances interfacial bonding and overall homogeneity within the mixture. As a result, RD-RAP mixtures demonstrate superior fracture energy compared with RC-RAP mixtures at equivalent RAP levels.

Figure 14
Results of SCB test: (a) fracture energy; (b) fracture toughness.

Fracture toughness emphasizes the critical ability to resist stress concentration at the crack tip during crack initiation, which is primarily governed by the ultimate load. The test results of fracture toughness are shown in Figure 14(b). With the incorporation of RAP, the stiffness of RHMA increased, thereby enhancing the ultimate load, and consequently the fracture toughness values also increased. This trend is opposite to that observed for Fracture energy. Although the addition of RAP improves the ultimate load of RHMA, once a crack is initiated, it becomes difficult to arrest crack propagation. As a result, cracks expand rapidly, the mixture fractures abruptly, and the energy required for the entire cracking process is substantially reduced. Hence, the risk of cracking in RHMA increases with RAP incorporation.

In addition, the SCB test results revealed that RD-RAP mixtures with 70% RAP exhibited comparable performance to RC-RAP mixtures with 50% RAP, indicating that refined decomposition offers greater advantages for the application of high-RAP mixtures.

5.6. Results of dynamic modulus test

The dynamic modulus test results of different RHMA types are shown in Figure 15. For RC-RAP mixtures, the dynamic modulus increased as RAP content rose from 0% to 50%, but declined when RAP reached 70%. This trend can be attributed to the higher pro-portion of aged asphalt, which enhances stiffness up to a point, while excessive RAP introduces more agglomerated particles and initial defects that limit further modulus growth. In contrast, RD-RAP mixtures exhibited a continuous increase in dynamic modulus with RAP content, maintaining an upward trend even at 70% RAP. Although the modulus of RD-RAP mixtures was slightly lower than that of RC-RAP mixtures at the same RAP level, this difference is likely due to reduced agglomeration and improved asphalt blending achieved by refined decomposition. Overall, while RC-RAP mixtures showed a modulus reduction beyond 50% RAP, RD-RAP mixtures sustained higher dynamic modulus at elevated RAP levels, confirming that refined decomposition enables more effective utilization of high RAP contents.

Figure 15
Results of dynamic modulus test.

6. CONCLUSIONS

This study compared the effects of RD and RC preprocessing methods on the performance of RHMA with different RAP contents, including optimum asphalt content, compaction characteristics, high-temperature rutting resistance, moisture stability, low-temperature cracking resistance and dynamic modulus. Based on the experimental results, the following conclusions can be drawn:

• (1)

  RD-RAP demonstrated superior Marshall performance over RC-RAP, with higher stability, lower air voids, and greater VFA. The OAC of RHMA increased with RAP content for both preprocessing methods, while RD-RAP mixtures consistently required 0.1%–0.3% less binder than RC-RAP mixtures at the same RAP level.

• (2)

  Compaction tests indicated that CEI was strongly influenced by RAP content, in-creasing with higher RAP levels. Compared with RC-RAP mixtures, RD-RAP mixtures exhibited 13.95%–31.58% lower CEI at the same RAP content. Refined decomposition can thus effectively mitigate the compaction difficulties associated with high RAP contents.

• (3)

  In terms of road performance, RAP incorporation enhanced the high-temperature stability of RHMA but adversely affected its moisture resistance and low-temperature cracking resistance. Compared with RC-RAP mixtures, RD-RAP mixtures exhibited slightly lower rutting resistance but achieved higher TSR and fracture energy. More-over, as RAP content increased to 50%–70%, RD-RAP mixtures maintained relatively high dynamic modulus, whereas RC-RAP mixtures showed a declining trend. At the same RAP content, RD-RAP mixtures exhibited superior road performance, allowing the permissible RAP content to be increased to 70%.

In summary, compared with roll crushing, refined decomposition not only ensured satisfactory high-temperature performance but also more effectively improved moisture stability, low-temperature cracking resistance, and compaction characteristics, with its advantages becoming more pronounced at high RAP contents (≥50%). This method pro-vides a feasible approach for enhancing the efficient utilization of RAP and holds significant potential for advancing sustainable road construction.

7. ACKNOWLEDGMENTS

This research was funded by the Transportation Department of Jiangxi Province (grant number: 2023Z0001, 2024YB047), Key Research and Development Program of Jiangxi Province (grant number: 20252BCG330027), Science and Technology Project (Youth Program) of the Jiangxi Provincial Department of Education (GJJ2500506).

8. BIBLIOGRAPHY

• ^[1] AL-SABAEEI, A.M., NAPIAH, M.B., SUTANTO, M.H., et al., “A systematic review of bio-asphalt for flexible pavement applications: coherent taxonomy, motivations, challenges and future directions”, Journal of Cleaner Production, v. 249, pp. 119357, 2020. doi: https://doi.org/10.1016/j.jclepro.2019.119357.
  » https://doi.org/10.1016/j.jclepro.2019.119357
• ^[2] ANTUNES, V., FREIRE, A.C., NEVES, J., “A review on the effect of RAP recycling on bituminous mixtures properties and the viability of multi-recycling”, Construction & Building Materials, v. 211, pp. 453–469, Jun. 2019. doi: https://doi.org/10.1016/j.conbuildmat.2019.03.258.
  » https://doi.org/10.1016/j.conbuildmat.2019.03.258
• ^[3] AI, X., PEI, Z., XU, K., et al., “RAP agglomeration and partial blending of recycled hot mix asphalt: A literature review”, Journal of Road Engineering, v. 5, n. 2, pp. 230–243, Jun. 2025. doi: https://doi.org/10.1016/j.jreng.2024.12.006.
  » https://doi.org/10.1016/j.jreng.2024.12.006
• ^[4] YU, D., LUO, J., YANG, J., et al., “Effect of waste oil rejuvenators on hot in-place recycled SMA pavement”, Case Studies in Construction Materials, v. 22, pp. e04813, 2025. doi: https://doi.org/10.1016/j.cscm.2025.e04813.
  » https://doi.org/10.1016/j.cscm.2025.e04813
• ^[5] ZHU, C., YANG, Y., ZHANG, K., et al., “Study on the road performance and compaction characteristics of fiber-reinforced high-RAP plant-mixed hot recycled asphalt mixtures”, Polymers, v. 16, n. 14, pp. 2016, Jul. 2024. doi: https://doi.org/10.3390/polym16142016. PubMed PMID: 39065333.
  » https://doi.org/10.3390/polym16142016
• ^[6] AI, X., PEI, Z., CAO, J., et al., “Environmental impact assessment of recycled HMA with RAP materials from the rotary decomposition process compared to virgin HMA and conventional recycled HMA”, Journal of Cleaner Production, v. 389, pp. 136078, Feb. 2023. doi: https://doi.org/10.1016/j.jclepro.2023.136078.
  » https://doi.org/10.1016/j.jclepro.2023.136078
• ^[7] SONG, W., HUANG, B., SHU, X., “Influence of warm-mix asphalt technology and rejuvenator on performance of asphalt mixtures containing 50% reclaimed asphalt pavement”, Journal of Cleaner Production, v. 192, pp. 191–198, Aug. 2018. doi: https://doi.org/10.1016/j.jclepro.2018.04.269.
  » https://doi.org/10.1016/j.jclepro.2018.04.269
• ^[8] FALCHETTO, D.A.C., MOON, D.K.H., KIM, D.H., “Evaluation of recycled asphalt mixture at low temperature using different analytical solutions”, Canadian Journal of Civil Engineering, v. 47, n. 7, pp. 801–811, 2020. doi: https://doi.org/10.1139/cjce-2019-0303.
  » https://doi.org/10.1139/cjce-2019-0303
• ^[9] AI, X., CAO, J., FENG, D., et al., “Performance evaluation of recycled asphalt mixtures with various percentages of RAP from the rotary decomposition process”, Construction & Building Materials, v. 321, pp. 126406, Feb. 2022. doi: https://doi.org/10.1016/j.conbuildmat.2022.126406.
  » https://doi.org/10.1016/j.conbuildmat.2022.126406
• ^[10] GAO, J., YAO, Y., SONG, L., et al., “Determining the maximum permissible content of recycled asphalt pavement stockpile in plant hot-mix recycled asphalt mixtures considering homogeneity: A case study in China”, Case Studies in Construction Materials, v. 16, pp. e00961, 2022. doi: https://doi.org/10.1016/j.cscm.2022.e00961.
  » https://doi.org/10.1016/j.cscm.2022.e00961
• ^[11] LIU, Q., YU, B., FALCHETTO, A.C., et al., “Characterization and molecular mechanism of the thermal-oxidative gradient aging behavior in asphalt films”, Measurement : Journal of the International Measurement Confederation, v. 199, pp. 111567, 2022. doi: https://doi.org/10.1016/j.measurement. 2022.111567.
  » https://doi.org/10.1016/j.measurement.2022.111567
• ^[12] LEI, G., PENG, L., “Investigation of optimum RAP contents in hot recycled asphalt mixture”, Huadong Jiaotong Daxue Xuebao, v. 35, n. 1, pp. 103–108, 2018. doi: https://doi.org/10.16749/j.cnki.jecjtu.2018.01.016.
  » https://doi.org/10.16749/j.cnki.jecjtu.2018.01.016
• ^[13] GUO, F., PEI, J., ZHANG, J., et al., “Study on adhesion property and moisture effect between SBS modified asphalt binder and aggregate using molecular dynamics simulation”, Materials, v. 15, n. 19, pp. 6912, 2022. doi: https://doi.org/10.3390/ma15196912. PubMed PMID: 36234252.
  » https://doi.org/10.3390/ma15196912
• ^[14] ZAUMANIS, M., OGA, J., HARITONOVS, V., “How to reduce reclaimed asphalt variability: A full-scale study”, Construction & Building Materials, v. 188, pp. 546–554, 2018. doi: https://doi.org/10.1016/j.conbuildmat.2018.08.137.
  » https://doi.org/10.1016/j.conbuildmat.2018.08.137
• ^[15] GAO, J., YANG, J., YU, D., et al., “Reducing the variability of multi-source reclaimed asphalt pavement materials: A practice in China”, Construction & Building Materials, v. 278, pp. 122389, Apr. 2021. doi: https://doi.org/10.1016/j.conbuildmat.2021.122389.
  » https://doi.org/10.1016/j.conbuildmat.2021.122389
• ^[16] CENTOFANTE, R., SPECHT, L.P., ALMEIDA JUNIOR, P.O.B.D., et al., “Avaliação do comportamento de misturas asfálticas recicladas a quente com inserção de material fresado”, Matéria (Rio de Janeiro), v. 23, n. 3, pp. e12178, 2018. doi: https://doi.org/10.1590/s1517-707620180003.0512.
  » https://doi.org/10.1590/s1517-707620180003.0512
• ^[17] ZAUMANIS, M., MALLICK, R.B., FRANK, R., “100% recycled hot mix asphalt: A review and analysis”, Resources, Conservation and Recycling, v. 92, pp. 230–245, 2014. doi: https://doi.org/10.1016/j.resconrec.2014.07.007.
  » https://doi.org/10.1016/j.resconrec.2014.07.007
• ^[18] QU, L., WANG, Y., WANG, L., et al., “Experimental evaluation of fatigue performance of recycled asphalt mixture using refined separation recycled aggregates”, Construction & Building Materials, v. 411, pp. 134786, Jan. 2024. doi: https://doi.org/10.1016/j.conbuildmat.2023.134786.
  » https://doi.org/10.1016/j.conbuildmat.2023.134786
• ^[19] FENG, D., CAO, J., GAO, L., et al., “Recent developments in asphalt-aggregate separation technology for reclaimed asphalt pavement”, Journal of Road Engineering, v. 2, n. 4, pp. 332–347, 2022. doi: https://doi.org/10.1016/j.jreng.2022.07.002.
  » https://doi.org/10.1016/j.jreng.2022.07.002
• ^[20] GE, D., YOU, Z., CHEN, S., et al., “The performance of asphalt binder with trichloroethylene: improving the efficiency of using reclaimed asphalt pavement”, Journal of Cleaner Production, v. 232, pp. 205–212, 2019. doi: https://doi.org/10.1016/j.jclepro.2019.05.164.
  » https://doi.org/10.1016/j.jclepro.2019.05.164
• ^[21] QIU, J., HUURMAN, M., DE BRUIN, B., et al., “Towards 90% warm re-use of porous asphalt using foaming technology”, Journal of Cleaner Production, v. 190, pp. 251–260, 2018. doi: https://doi.org/ 10.1016/j.jclepro.2018.04.086.
  » https://doi.org/10.1016/j.jclepro.2018.04.086
• ^[22] YU, X., TANG, W., LI, N., et al., “Refined decomposition: A new separation method for RAP materials and its effect on aggregate properties”, Construction & Building Materials, v. 358, pp. 129452, Dec. 2022. doi: https://doi.org/10.1016/j.conbuildmat.2022.129452.
  » https://doi.org/10.1016/j.conbuildmat.2022.129452
• ^[23] TANG, W., LI, N., YU, X., et al., “Aggregate uniformity of recycled asphalt mixtures with RAP from refined decomposition process: Comparison with routine crushing method”, Construction & Building Materials, v. 442, pp. 137665, Jul. 2024. doi: https://doi.org/10.1016/j.conbuildmat.2024.137665.
  » https://doi.org/10.1016/j.conbuildmat.2024.137665
• ^[24] WANG, Y., DOU, G., WANG, S., et al., “The effect of refined separation on the properties of reclaimed asphalt pavement materials”, Buildings, v. 14, n. 6, pp. 1608, Jun. 2024. doi: https://doi.org/10.3390/buildings14061608.
  » https://doi.org/10.3390/buildings14061608
• ^[25] MINISTRY OF TRANSPORT OF THE PEOPLE’S REPUBLIC OF CHINA, JTG/5521-2019: Technical Specifications for Highway Asphalt Pavement Recycling, Beijing, Ministry of Transport of the People’s Republic of China, 2019.
• ^[26] ZOU, X., LI, N., WU, W., et al., “Study on Effect of Fine Separation on RAP Material Properties”, Journal of Dalian Jiaotong University, v. 44, n. 6, pp. 71–76, 2023. doi: https://doi.org/10.13291/j.cnki.djdxac.2023.06.012.
  » https://doi.org/10.13291/j.cnki.djdxac.2023.06.012
• ^[27] MINISTRY OF TRANSPORT OF THE PEOPLE’S REPUBLIC OF CHINA, JTG E42-2005: Test Methods of Aggregate for Highway Engineering, Beijing, Ministry of Transport of the People’s Republic of China, 2005.
• ^[28] MINISTRY OF TRANSPORT OF THE PEOPLE’S REPUBLIC OF CHINA, JTG E20-2011: Standard Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering, Beijing, Ministry of Transport of the People’s Republic of China, 2011.
• ^[29] YIN, P., PAN, B., “Formulation of a new rejuvenator and its regenerative effect on aged asphalt”, Journal of Materials in Civil Engineering, v. 34, n. 12, pp. 04022341, 2022. doi: https://doi.org/10.1061/(ASCE)MT.1943-5533.0004509.
  » https://doi.org/10.1061/(ASCE)MT.1943-5533.0004509
• ^[30] CHEN, M., LENG, B., WU, S., et al., “Physical, chemical and rheological properties of waste edible vegetable oil rejuvenated asphalt binders”, Construction & Building Materials, v. 66, pp. 286–298, 2014. doi: https://doi.org/10.1016/j.conbuildmat.2014.05.033.
  » https://doi.org/10.1016/j.conbuildmat.2014.05.033
• ^[31] MINISTRY OF TRANSPORT OF THE PEOPLE’S REPUBLIC OF CHINA, JTG F40-2004: Technical Specification for Construction of Highway Asphalt Pavements, Beijing, Ministry of Transport of the People’s Republic of China, 2004.
• ^[32] HAN, S., CHENG, X., LIU, Y., et al., “Laboratory performance of hot mix asphalt with high Reclaimed Asphalt Pavement (RAP) and Fine Reclaimed Asphalt Pavement (FRAP) content”, Materials, v. 12, n. 16, pp. 2536, 2019. doi: https://doi.org/10.3390/ma12162536. PubMed PMID: 31395829.
  » https://doi.org/10.3390/ma12162536
• ^[33] MA, X., LENG, Z., WANG, L., et al., “Effect of reclaimed asphalt pavement heating temperature on the compact ability of recycled hot mix asphalt”, Materials, v. 13, n. 16, pp. 3621, 2020. doi: https://doi.org/10.3390/ma13163621. PubMed PMID: 32824382.
  » https://doi.org/10.3390/ma13163621
• ^[34] FAN, X., CAO, L., XIONG, Y., et al., “Research on recycled asphalt mixture with high RAP dosage based on separation of oil and stone”, Journal of Central South University, v. 56, n. 3, pp. 1080–1090, 2025. doi: https://doi.org/10.11817/j.issn.1672-7207.2025.03.022.
  » https://doi.org/10.11817/j.issn.1672-7207.2025.03.022
• ^[35] LI, N., ZHAN, H., LI, H., et al., “Investigation of recycled asphalt pavement materials treated by refined crushing and screening process: comparison with conventional methods”, Journal of Materials in Civil Engineering, v. 36, n. 2, pp. 04023579, Feb. 2024. doi: https://doi.org/10.1061/JMCEE7.MTENG-16734.
  » https://doi.org/10.1061/JMCEE7.MTENG-16734
• ^[36] WANG, P., CHEN, J., WANG, J., et al., “Evaluation of the refined decomposition effect of reclaimed asphalt pavement materials”, Buildings, v. 13, n. 9, pp. 2240, Sep. 2023. doi: https://doi.org/10.3390/buildings13092240.
  » https://doi.org/10.3390/buildings13092240

Publication Dates

• Publication in this collection
  16 Mar 2026
• Date of issue
  2026

History

• Received
  27 Sept 2025
• Accepted
  27 Jan 2026