Study on performance of recycled asphalt mixture based on blending state analysis of virgin and aged asphalt

ABSTRACT The blending state of virgin and aged asphalt is one of the key factors that influence the cracking resistance and fatigue performance of recycled asphalt mixtures. In this paper, the complex modulus of the virgin and aged double-layered asphalt samples was measured by DSR experiments, and the degree of blending (DOB) was calculated based on the complex modulus. The cracking resistance and fatigue performance of the mixture prepared under different blending conditions were tested and related to the DOB to obtain a prediction model. Results showed that the DOB and the complex modulus G* increased with blending time and temperature. The blending effect of modified asphalt was not as good as 70-pen asphalt. The contribution of blending time and temperature to DOB was 40% and 50%, respectively. The influence of DOB on cracking resistance and fatigue performance of recycled asphalt mixtures was consistent with DSR test. Good correlations between DOB, cracking resistance and fatigue performance were observed, which were 79.14% and 82.82%, respectively. Based on the DOB model, optimal blending temperature and blending time could be found according to the softening point difference between virgin and aged asphalt.


Introduction
Recycled asphalt mixture can be regarded as a mixture composed of virgin and reused materials, including the interface between virgin and aged asphalts.The blending state of virgin and aged asphalt directly influences the mechanical properties of the aggregate contact interface, affects the internal stressstrain transfer of the mixture, and then determines the macroscopic mechanical properties of the mixture.Therefore, fully understanding the blending behaviour characteristics of virgin and aged asphalts in the recycled mixture and improving the mechanical properties of the interface blending zone of virgin and aged asphalt in the recycled mixture is critical in effectively improving the service life.It is also the key factor in improving the blending rate of RAP.

Literature review
According to the existing engineering practice, the recycled asphalt pavement (RAP) content of hot mix recycled asphalt mixture should be controlled at 20% ∼ 30%.(Huang et al. 2001, Zha et al. 2013, Ding et al. 2019, Kaseer et al. 2019, Zhu et al. 2020, Pinheiro et al. 2022).When rejuvenators and balanced mix design were used, 100% RAP content pavements could be achieved (Zaumanis et al. 2014 and2019).However, the fatigue durability of recycled asphalt mixture is still compromised at high RAP content.Therefore, in practical engineering applications, the deficiency of road performance of recycled pavement with high RAP content directly limits the application and development of hot recycled mixture technology (Ma et al. 2020b, Yang et al. 2021, Magar et al. 2022).
Studies have shown that the key to regeneration technology lies in the interface between RAP surface aging and virgin asphalts (Karlsson and Isacsson 2003, Ma et al. 2015, Navaro et al. 2012, Zhao et al. 2015, Chen et al. 2020).In the recycling process, the blending state of virgin and aged asphalt is the key factor affecting the road performance and pavement service life of recycled asphalt mixture.From the macro, meso and micro levels, many researchers used a variety of technical means and evaluation methods to study the blending process of virgin and aged asphalts (Rad et al. 2014, Shi et al. 2019, Guo et al. 2020).
In terms of macro scale, Qin et al. (2015) used indirect tensile modulus, splitting strength and immersion splitting strength as evaluation indexes to ascertain the performance of recycled mixture through the difference of mechanical properties between virgin and aged mixtures to indirectly determine the blending ratio of virgin and aged asphalts.Cavalli et al. (2017) used electron microscopy and image analysis techniques to measuring the binder film residues on black rock in mixtures.Results show that RAP aggregates have a layered structure of RAP binder and virgin bitumen and the higher the mixing temperature, the thinner the RAP binder film.Peng (2019) found that with the increase of the blending degree of virgin and aged asphalts, the free asphalt content in RAP increases, and the cementation between asphalt and aggregate increases, whereas the dynamic stability of recycled asphalt mixture significantly decreases.Further, the water stability, low-temperature performance and fatigue resistance are improved.Hao et al. (2021) showed that the recycled mixture with the complete blending of virgin and aged asphalt has poorer high-temperature performance and better low-temperature performance, cracking resistance and water stability than the conventional mixing method.When the RAP content is more than 30%, the influence of the blending of virgin and aged asphalts should be fully considered in the hot recycled asphalt mixture.Zhao and Chen (2011) confirmed that there is a good correlation between the blending degree of virgin and aged asphalts and the cracking resistance of recycled asphalt mixture.Oliveira et al. (2022) evaluated the fatigue life of hot recycled asphalt mixtures with different RAP incorporation levels; the results showed that mixtures with higher incorporations tended to have a shorter fatigue life.
In terms of micro-level and micro-scale, Zhao (2016) used Fourier transform infrared spectroscopy (FTIR) to explore the effects of mixing temperature, mixing time and RAP content on the interfacial blending degree of virgin and aged asphalt.It was shown that the phenomenon of non-uniform blending of virgin and aged asphalt is common, and the blending degree of the outer asphalt film of RAP and the virgin asphalt is greater than that of the inner asphalt film.Lei et al. (2021) identified the difference in the mobilisation degree of RAP asphalt binder under three mixing techniques.The results indicated that the softer fresh asphalt binder was more capable of facilitating the mineral aggregates to form a stable skeleton structure.Navaro et al. (2012) marked the regeneration agent with pigment, microscopically observed the blending state of virgin and aged asphalt, and characterised the spatial distribution of virgin and aged asphalt and regeneration agent through the grey value distribution of the picture.Rinaldini et al. (2014) used the dynamic shear rheometer (DSR), electron microscopy and computer tomography (CT) to investigate the blending of reclaimed asphalt with virgin materials.Results show that blending was not homogeneous throughout the sample and mapping the binder using tracer elements would be a useful way for observation of mobilisation degree of RAP.Bowers et al. (2014) quantitatively analysed the interfacial blending behaviour of virgin and aged asphalt based on gel chromatography (GPC) and FTIR, and demonstrated that only partial blending occurred in the interface area of virgin and aged asphalts.Zhao et al. (2015) studied the interfacial blending behaviour of waste asphalt mixture using GPC and atomic force microscope (AFM).It is found that the diffusion depth of virgin and aged double-layer asphalt samples at its interface reached 25 μm.In addition, some researchers (Al-Qadi et al. 2009, Hu et al. 2022, Sun et al. 2022, Yu et al. 2022) combined the scanning electron microscope (SEM) and energy dispersive spectroscopy (EDS) to observe the blending state of virgin and aged asphalts.Wu et al. (2022) studied cohesive and adhesive properties of emulsified cold recycled mixture with molecular dynamics simulations.Results showed that modifiers and additives were necessary to enhance interfacial adhesion.
In order to have a better understanding of the blending phenomena between virgin and aged asphalt, concepts of the Degree of Blending (DOB) and the Degree of binder Availability (DOAv) were put forward into practical suggestions in mix design procedures (Davide et al. 2020).Solaimanian et al. (2021) investigated the DOB through the Superpave volumetric mix design.And it was estimated that approximately 75% residual RAP binder interacted with the virgin binder in terms of blending and contribution to the final mix.Ashtiani et al. (2022) blended aged asphalt with different proportions of a virgin binder, and developed a locally calibrated Hirsch model to estimate master curves for the asphalt mixtures from the corresponding binder-blend data.Results showed that degree of blending occurred at 50% to 60% for RAP and at 15% to 30% for RAS.Ma et al., (2020a) analysed the rheological and aging properties of asphalt binders blended with different rejuvenator application procedures through the dynamic shear rheometer (DSR) test.Results revealed that the rejuvenator application procedures significantly influenced the workability, rheological property and aging effect of the asphalt binder.
The literature review shows that performance testing in macro sacle could be used to analyse the blending state of virgin and aged asphalt, such as DSR test for mixed binder and cracking resistance and fatigue performance tests for mixtures.Although there were research findings about the quantitative characterisation of the blending behaviour of virgin and aged asphalts in micro scale, the relationship between the blending state and the mixture performance also needs further exploration.

Aims and objectives
In most cases, an assumption of 100% blending of virgin and aged asphalt is used in hot recycled mixture design, and the blending state of virgin and aged asphalts is not fully considered.As a result, the performance of recycled mixture differs from the theoretical hypothesis, leading to a poor recycling effect.It is worth noting that there is still no consensus on how much aged asphalt is involved in recycling, how the blending state influences the macro-performance of the recycled mixture, and which factors affecting the blending state that are controllable in practical engineering applications.As a matter of that, the recycling process parameters still need to be adjusted according to the property of virgin and recycled materials used.
Based on the above background, the dynamic shear tests on virgin and aged asphalt double-layer samples were carried out in this paper.The rheological characteristics of double-layer samples under multiple factors were tested, and the Degree of Blending (DOB) was calculated.On this basis, the relationship between the DOB and the performance of recycled mixture is evaluated, and a correlation model is established.The results of this research could help to enhance the performance of recycled mixture and provide a simple reference for engineering application.

Raw materials
The virgin asphalt used in this study is 70-pen asphalt and SBS-modified asphalt (with 4% SBS content).The mineral aggregates used are diabase, and mineral powder is limestone.Referring to the relevant research (Li et al. 2008, Yao et al. 2019), the Rolling Thin Film Oven Test (RTFOT) is used to prepare aged asphalt.Two aging conditions are set, for the short-term aging and long-term aging of asphalt, the heating time of RTFOT tests are 90 and 360 min at 163C, respectively.Therefore, six asphalts with different properties are used in this test, which are 70-pen asphalt, 70-short term, 70-long term, modified asphalt, modified-short term and modified-long term, respectively.
The test results of technical indexes of virgin asphalt, aged asphalt and aggregate are shown in Tables 1 and 2.

Asphalt mixture design
The mixture gradation used in this study is AC-13, the oilstone ratio is 4.6%, and the gradation composition is shown in Table 3.
In order to reduce the influence of RAP variability on the macroscopic performance test of recycled asphalt mixture (Zhang et al. 2010, Zhang 2013, Zhang and Qi 2016), 'RAP' is prepared by mixing the aged asphalt with the virgin aggregates.The prepared short-term aging and long-term aging asphalt are also used for the preparation of man-made RAP.In addition, the amount of RAP simulated in this study is 50%.

Methods and approach
In this study, the DSR experiments are carried out to obtain the complex modulus G* of the virgin and aged double-layered asphalt samples.The DOB is calculated based on the DSR test results.Then, the cracking resistance and fatigue performance tests of the mixture prepared under different blending conditions are tested.The flow chart of this study is showed in Figure 1.

DSR test and DOB calculation
The gap of the DSR test is set to 2 mm, the diameter of the rotor is 25 mm.Therefore, 2 mm thick specimen are moulded.The preparation process of virgin and aged asphalt doublelayer samples is, the first step, virgin and aged asphalt samples with thickness of 1 mm are moulded.Then, the virgin asphalt sample is placed on the top of aged asphalt sample to form a 2 mm thick double-layer sample.As shown in Figure 2.
In order to fabricate specimens under different blending conditions, the virgin and aged asphalt double-layer samples are kept in the mould and placed in the oven at different temperatures (90, 120, 150°C) for different durations (10, 30, 60, 90, 150 min).Finally, the double-layer specimens with different DOB for DSR test are obtained.
Considering the difference in performance between virgin and aged asphalts, the viscoelastic properties of double-layer samples depend on the blending state of virgin and aged asphalts.As a commonly used parameter, the complex modulus G* has the advantages of high sensitivity and is easy to measure.Therefore, complex modulus G* is selected in this study as to calculate the DOB of the samples.The DSR test parameters are summarised in Table 4.
Four combinations of virgin and aged asphalt are used in this study, which are Three specimens for each combination are moulded.For the virgin and aged asphalt double-layer samples in this study, the blending process can be explained in Figure 3.
According to Hettiarachchi's study (Hettiarachchi et al. 2020), the DOB is used to characterise the blending degree of virgin and aged asphalt.The complex modulus G* in the zero-blending state (Figure 3(a)) and complete-blending state (Figure 3(d)) are measured.The test started immediately after double-layer specimen fabricated is used to represent the zero-blending state, and 2 mm thick specimen prepared with fully mixing 50% virgin asphalt and 50% aged asphalt is used to represent the completely blending state.Then, combined with the complex modulus G* of virgin and aged asphalt double-layer specimens under different blending conditions, the DOB could be calculated by linear interpolation according to Equation (1).
where G * zero and G * complete are the complex modulus of doublelayer specimens with zero-blending state and completelyblending state, respectively.

Preparation of mixtures
The recycled asphalt mixture is prepared in the following way.In the first step, the aged asphalt and virgin aggregate are mixed to get man-made RAP.Secondly, virgin asphalt, virgin aggregate and mineral powder are added and mixed with manmade RAP.Then, the mixture is put into the oven, and kept at different temperatures for different durations to prepare recycled asphalt mixtures with different DOB.
According to the blending conditions set in the preparation of virgin and aged asphalt double-layer samples, the blending condition of splitting test is set as follows: blank control group (specimens are moulded immediately after mixing complete) and test group.In the test group, before moulded, mixtures are put in the 120 °C oven for 30 min, 120 °C oven for 60 min, 120 °C oven for 90 min, 150 °C oven for 30 min, 150 °C oven for 60 min, 150 °C oven for 90 min, respectively.Also, the blending condition of the fatigue test is set as blank control group and test group.In the test group, mixtures are put in 120 °C oven for 30 min, 150 °C oven for 30 min and 150 °C oven for 60 min before moulded, respectively.
At last, the recycled mixture is kept at 160 °C for 10 min in the oven to ensure that the mixture is moulded at the same temperature.
Reasearches have proved that recycled asphalt mixtures are sensitive to cracking and fatigue failure (Stephens et al. 2001, Al-Qadi 2007, Mannan et al. 2015, Sabouri et al. 2015, Daniel et al. 2019).The splitting and fatigue performance tests are adopted to analysed the influence of interfacial blending behaviour of virgin and aged asphalt on recycled asphalt mixture.Marshall specimens are moulded for the splitting test.The Marshall specimen is double-sided and compacted for 75 times at 160°C.Beam specimens for four-point bending fatigue test are fabricated using wheel rolling moulding machine and a double-sided saw for cutting.

Performance test of mixtures
According to the Standard Test Methods of Bitumen and Bituminous Mixtures for Highway Engineering (JTG E20-2011),   According to the Standard (JTG E20-2011), the fatigue life is defined as the loading repetitions when the modulus of specimen drops to 50% of the initial value.The equipment used in this study is COOPER NU-14, equipped with an environment chamber.The fatigue specimen is a 380 mm*63 mm*50 mm beam.The test temperature is 15 °C.The strain control mode with 200 microstrains is adopted in this study, and the test frequency is 10 Hz.Under each blending condition, three specimens for fatigue test are moulded.

Complex modulus G* and DOB
The complex modulus G* of the double-layer specimens was obatained by DSR test.The results of specimens with zeroblending state and completely-blending state are shown in Table 5.And complex modulus G* under different blending conditions are shown in Figure 4.
According to Equation (1), the DOB of all double-layer specimens under different blending conditions are calculated.The statistical results are shown in Figure 5.
It is worth noting that DOB is greater than 100% in some blending conditions.For example, 70 + 70ST combination under 150 °C + 150 min condiction and 70 + 70LT under 150 °C + 90 min condition.This is due to the aging of both virgin and aged asphalt caused by a long holding time at high temperature, which greatly changes the performance of asphalt.

Effect of temperature
As known from Figure 4 and Figure 5, under the premise of the same blending time, with the increase of blending temperature, the complex modulus G* of double-layer specimens increases, and the DOB increases.The state of aged asphalt influences the DOB value.Under the blending condition of 90°C+150 min, the DOB of 70 + 70ST combination is 35.49%, while that of 70 + 70LT combination is merely 28.19%, indicating that the blending of virgin and aged asphalt needs more energy when asphalt is severely aged.
Under the same blending time (60 min), the DOB of 70 + 70ST combination increased by 36.18%, and the DOB of 70 + 70LT combination increased by 28.32% when the temperature increases from 90°C to 120°C.And the DOB increases by 25.32% and 14.73%, respectively, when the temperature increases from 120°C to 150°C.There are also similar results for both M+MST and M+MLT combinations.The results show that increasing the temperature can greatly improve the DOB of virgin and aged asphalts.

Influence of time
With the increase of blending time, the complex modulus G* and DOB increased.Based on the blending condition set, the differences in the DOB increment of blending time in the pre-60 min are compared, and the results are shown in Table 6.
It can be seen from Table 6 that the blending effect is different at different blending time.The DOB increment of pre-30 min is greater than that of 30-60 min for all four combinations, indicating that the blending efficiency of virgin and aged asphalt in the early stage is better than that in the later stages.

Influence of aged degree
The DOB ratio of short-term combination to long-term combination is calculated, as shown in Table 7.
It can be seen that nearly all of the DOB ratios are greater than 1, which indicates that the DOB of short-term combination is higher than that of long-term combination under the same blending condition.Compared with the long-term combination, the aging degree of asphalt in the short-term combination is relatively less, and the asphalt components of virgin and aged asphalt is relatively similar.Therefore, virgin and aged asphalt in the short-term combination are easier to blend together to form a uniform and stable system.

Influence of asphalt types
The increase of the G* from 90°C to 150°C is calculated at different blending time.The results are tabulated in Table 8.
According to Table 8, the complex modulus of 70 + 70ST combination increases by 23.11% at 30 min, while that of M + MST combination increases by 1.58%.The complex modulus of the 70 + 70LT combination increases by 58.26% at 30 min, while that of the M + MLT combination increases by 3.05%.There are similar results for the other blending time.These results show that it is more difficult to improve the viscoelastic performance of aged modified asphalt than 70pen asphalt.The reason may be that the macromolecules and network structure of SBS modifiers increases the energy required for blending between vrigin and aged modified asphalt.

Parameter selection
The DOB of partial combination and partial blending conditions are selected to study the relationship between blending time, blending temperature and DOB.The result is shown in Figure 6.
According to Figure 6, the linear correlation coefficients between blending time, temperature and DOB are more than 80%.Therefore, blending time and temperature could be used as parameters in the DOB prediction model.Secondly, the aging degree and the type of asphalt also influence the blending effect.In this study, the influence of aging  degree and asphalt type is attributed to the performance difference between virgin and aged asphalts.The relationship between the difference of softening point and DOB is shown in Figure 7.
As shown in Figure 7, there is a logarithmic correlation between the difference of softening point and DOB, and the correlation coefficient is more than 90%.Therefore, the difference of softening point between virgin and aged asphalt could be used as a parameter of the DOB prediction model.
In summary, the virgin and aged asphalts could be regarded as an energy system.The blending process is essential that the system absorbs energy from the outside, which promotes the blend between the virgin and aged asphalt components.Therefore, this energy system is significantly sensitive to blending time, blending temperature and the difference of softening points.The following equation is established based on the regression method.
where T is blending temperature, °C; t is blending time, min; DS is the difference of softening point between virgin and aged asphalt, °C; a, b, c are model coefficients.

Model calibration
Based on the DOB results under different blending conditions, the above models is fitted and regressed by the data statistics software OriginLab.The calibrated DOB prediction model is as follows: The fitness of the calibrated prediction model is tested.The goodness-of-fit test table and variance analysis of the model are shown in Table 9.According to Table 9, under the premise of knowing the blending temperature, blending time and softening point of virgin and aged asphalt, the DOB can be anticipatorily calculated.The model reflects the comprehensive influence of the external environment and material properties on the blending effect.As can be seen from the model parameters 0.508 and 0.399, the contribution of blending temperature and blending time to the degree of blending is 50% and 40%, respectively, indicating that the effect of temperature on blending is more significant.The '−17.23 ln DS' item in the model reflects that a smaller performance difference between the virgin and aged asphalt is beneficial to the blending effect.Furthermore, the smaller the softening point difference, the smaller the concentration gradient of the virgin and aged asphalt, the more uniform the materials in the interface area, and the closer the physical properties (density, hardness, solubility and thermal conductivity) of the virgin and aged asphalt at the interface.The continuity of physical properties provides a good basis for the blending effect in the interface area.10.According to Table 10, the difference of technical index between blending conditions is small.Therefore, it can be considered that keeping the mixtures at 160 °C for 10 min in the oven before moulded is beneficial for the consistency of the specimen.

Results of splitting test
The splitting tests under different blending conditions are carried out, and the test results are shown in Table 11.
According to Table 11, the splitting strength of the control group is lower than the test group, the splitting strength of the control group is 0.8856 MPa, and the maximum splitting strength of the test group is 1.5238 MPa.With the increase in blending time and the blending temperature, the splitting strength increases gradually.Under 120°C, the splitting strengths increase by 8.99%, 26.87% and 42.84% with 30, 60 and 120 min blending time, respectively, compared to the control group.When the blending temperature rises to 150°C, the splitting strengths increase by 20.33%, 44.94% and 72.06% with 30, 60 and 120 min blending time, respectively.This result indicates that a higher blending temperature is more beneficial to improve the performance of recycled asphalt mixtures, which is consistent with the results of DSR test.The better the blending effect of the virgin and aged asphalts in recycled asphalt mixture, the better the tensile performance of the recycled mixture.

Results of fatigue test
The results of the four-point bending fatigue test are shown in Table 12.
According to Table 12, it is clear that with the increase of blending time and temperature, the DOB of virgin and aged asphalt in recycled asphalt mixture increase, and the fatigue performance of recycled asphalt mixture improves.Compared with the control group, the fatigue life of the mixture under the blending condition of 120°C+30 min, 120°C+60 min and 150°C +60 min are increased by 11.75%, 29.03% and 79.72%, respectively.Although the test temperature is the same, under the same blending condition, the increase of performance index of the fatigue test is not consistent with the splitting test.The increased ratio of fatigue life higher than that of the splitting strength, which indicates that test methods have different sensitivity to the DOB.Nevertheless, increasing the DOB is beneficial to improve the uniformity of asphalt film inside recycled asphalt mixture and effectively delays the development of micro-cracks caused by fatigue loading in the interfacial blending zone.

Correlation analysis
The DOB obtained according to the test results of DSR for asphalt and performance of recycled asphalt mixtures is adopted in the correlation analysis between DOB, splitting strength and fatigue life, as shown in Figure 8.
As shown in Figure 8, the splitting strength and fatigue life increase with the increase of DOB.The linear correlation coefficients between DOB, splitting strength and fatigue life  where S SP is the splitting strength of recycled asphalt mixture, MPa; and N is the fatigue life of mixture; T is blending temperature, °C; t is blending time, min; DS is the difference of softening point between virgin and aged asphalt, °C.
As long as virgin asphalt is determined to be used in practical engineering, difference of softening points between virgin and aged asphalt can be calculated, and the blending temperature T and blending time t can be recommended according to the preset mixture performance requirements.

Conclusions
In this study, the DOB of double-layer specimens is calculated based on the DSR test, and the cracking resistance and fatigue performance of recycled asphalt mixture are related to the DOB to obtain a prediction model.The main conclusions of this study are shown below.
(1) Increasing blending time and temperature are able to effectively improve the DOB.The DOB of short-term combination is higher than that of long-term combination under the same blending condition, indicating that the blending needs more energy when asphalt is severely aged.The blending effect of modified asphalt is not as good as 70-pen asphalt.(2) A higher blending temperature is more beneficial to improve the cracking resistance and fatigue performance of recycled asphalt mixtures, which is consistent with the results of DSR test.
(3) The difference of softening point could be taken as a parameter in the DOB prediction model.According to the DOB prediction model, the contribution rates of blending time and temperature are 40% and 50%, respectively.(4) There are good correlations between DOB, cracking resistance and fatigue performance, which are 79.14% and 82.82%, respectively.Increasing the DOB is beneficial to improve the uniformity of asphalt film inside recycled asphalt mixture and, therefore, effectively delays the development of micro-damage in the interfacial blending zone.
However, these conclusions are based on limited experimental data.It is still imperative to discuss the mechanical properties of the mixture under different blending conditions and added regeneration agents.Future work will focus on solving the problem of compatible permeability of regenerant in the complex blending interface.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 2 .
Figure 2. Preparation of virgin and aged asphalt double-layer samples.

Figure 3 .
Figure 3. Blending process of virgin and aged asphalt.

4. 4 .
Performance test results of recycled asphalt mixture 4.4.1.Results of Marshall test The dosage of man-made RAP is 50% in this study.The test results of the technical index of the Marshall specimens used in the splitting test are shown in Table

Figure 7 .
Figure 7. Correlation between difference of softening point and DOB.

Funding
This work was supported by National Natural Science Foundation of China [grant number 51708144]; Guangzhou Municipal Science and Technology Project [grant number 202102010480].

Figure 8 .
Figure 8. Correlation between (a) DOB and splitting strength, and (b) DOB and fatigue life.

Table 1 .
Technical indexes of asphalt.

Table 2 .
Technical indexes of aggregates.

Table 3 .
Target gradation.Figure 1. Flow chart of the research approach.

Table 4 .
Setting of the DSR test.the 63.5 mm in height, 101 mm in diameter Marshall specimens are used in the splitting test.The loading rate of the test is 50 mm/min, and the test temperature is 15 °C.Under each blending condition, four specimens for the splitting test are moulded.

Table 5 .
Complex modulus of zero-blending state and completely blending state (Pa).

Table 6 .
Calculated results of DOB increments in different durations (%).

Table 7 .
DOB ratio under different aging degrees.

Table 8 .
Calculation results of G* growth rate (%).Correlation between temperature, blending time and DOB.

Table 9 .
Fitting result of DOB model.

Table 10 .
Technical index of Marshall specimens under different blending conditions.

Table 11 .
Results of splitting strength under different blending conditions (MPa).