3.1. Asphalt Viscosity Evaluation Test Results and Analysis
The viscosities of the asphalts at each test temperature are listed in
Table 3. Each asphalt sample was tested 3 times in parallel. The test results satisfy the allowable error of the repeatability test being 3.5% of the average value.
The viscosities of the virgin and reclaimed asphalt binders decreased gradually with increasing temperature, independent of whether they were aged or not. The viscosity of the reclaimed asphalt was significantly greater than that of the virgin asphalt, while the viscosity of the short-term aged asphalt was significantly greater than that of the virgin reclaimed asphalt.
To evaluate the influence of the source and admixture of the aged asphalt on the viscosity of the reclaimed asphalt, the relative viscosity Δ
η was used for the evaluation, which mainly characterizes the influence of the relative viscosity of the reclaimed asphalt on its viscosity for every 1% increase in the content of aged asphalt, calculated as Equation (1):
where Δ
η is the dimensionless viscosity of the reclaimed asphalt,
ηmix.r is the relative viscosity of the reclaimed asphalt,
ηmix.r =
ηmix/
ηnew,
ηmix is the viscosity of the reclaimed asphalt (Pa·s),
ηnew is the viscosity of the virgin asphalt (Pa·s),
ηnew.r is the relative viscosity of the virgin asphalt,
ηnew.r = 1, and
x is the amount of aged asphalt blending.
The variance results for the Δ
η and viscosity
η values of the reclaimed asphalt samples at different test temperatures and under different aging conditions are listed in
Table 4. Usually, the significance level α = 0.05.
As can be seen from
Table 4, the statistical probability
p-value of
η is less than 0.05 only for the test temperature and aging conditions, indicating that there is no significant difference in the effects of virgin asphalt and aged asphalt on the viscosity of reclaimed asphalt based on the η index. For Δ
η, the
p-values for all four influencing factors are less than 0.05, indicating that assessing the viscoelasticity of reclaimed asphalt with Δ
η can identify the differences in these four factors. Therefore, the high-temperature rheology of the reclaimed asphalt is better assessed using Δ
η than the viscosity index.
The variation in relative viscosity Δ
η of the reclaimed asphalt versus temperature is shown in
Figure 1. As can be seen from
Figure 1, the Δ
η of the reclaimed asphalt gradually decreases with the increase in temperature, and after short-term aging the Δ
η also gradually decreases. When the temperature is 135 °C, the Δ
η values of
A15
W,
B15
W, and
C15
W are 2.46, 2.11, and 1.75, respectively, indicating that the relative viscosity increases by 2.46, 2.11, and 1.75 for each 1% increase in the content of aged asphalt under this test condition. However, under the same temperature conditions, the Δ
η of B30
W is 2.63, which is not the same as that of B15
W, indicating that the viscosity of the reclaimed asphalt produces inconsistent changes under different aged asphalt admixtures, even if the aged asphalt is the same. Moreover, the different reclaimed asphalts at different test temperatures and different aging conditions produced similar test results as described above.
The regression analysis results show that the reclaimed asphalt Δ
η has a good linear relationship with the test temperature (T). The regression equations are all Δ
η = a
T + b (a and b are the fitting parameters), as shown in Equations (2) and (3).
From Equations (2) and (3), it can be seen that the effects of virgin asphalt and aged asphalt on the high-temperature rheology of the reclaimed asphalt can be characterized by the slope and intercept in the linear relationship equation. For example, the slope of B30W is −0.013, which is 23% and 33% smaller than for A30W and C30W, respectively. In addition, the Δη of the reclaimed asphalt is always linearly related to the temperature, independent of the source, the admixture of the aged asphalt, or whether it undergoes short-term aging.
3.2. Asphalt Rutting Factor Test Results and Analysis
The complex modulus G* can describe the ability of the asphalt to resist deformation, and the δ can reflect the proportional relationship between the elastic and viscous parts of the asphalt. Generally speaking, the larger δ is, the more viscous the asphalt is. From
Figure 2,
Figure 3 and
Figure 4, it can be seen that as the temperature increases, G* decreases and δ increases, and the regularity is not related to the source of the aged asphalt, the admixture of the aged asphalt, or whether it has been aged. The G* of the reclaimed asphalt is greater than that of the virgin asphalt under each test temperature condition, and the δ of the reclaimed asphalt is less than that of the virgin asphalt, whereby the lower the temperature the more significant the result. In addition, the aged asphalt with the higher admixture content has a larger G* and smaller δ.
Asphalt under the long-term coupling effects of heat, oxygen, light, water, and load will experience serious aging, which will be manifested in the components as a decrease in aromatic content, an increase in asphalt content, and a macroscopic increase in hardness. From a viscoelastic point of view, the viscosity of asphalt decreases, the elasticity increases, and the asphalt changes from the sol–gel state to the gel state, which leads to a higher G* and lower δ.
Adding a different proportion of aged asphalt to the virgin asphalt can improve the high-temperature performance of recycled asphalt; that is, the ability of the asphalt to resist high-temperature deformation. This is manifested as an increase in G* and a decrease in δ. According to the changes of G* and δ, it is considered that adding virgin asphalt to the aged asphalt can restore the rheological properties of the aged asphalt mixture.
3.3. Asphalt Rutting Factor Evaluation Test Results and Analysis
The road rutting is the irrecoverable deformation of asphalt pavement under the coupling effect of load and high temperature, which can be evaluated by using the rutting factor (G*/sinδ). The variation curves of the G*/sinδ with temperature for the virgin and reclaimed asphalts are shown in
Figure 5.
As shown in
Figure 5, the G*/sinδ values of the virgin and reclaimed asphalts gradually decreased with the increase in temperature. The G*/sinδ of the reclaimed asphalt was larger than that of the virgin asphalt. The nonlinear regression analysis showed that the G*/sinδ had a good exponential relationship with the temperature, and the correlation coefficients were all above 0.90.
In order to evaluate the effects of the source and admixture of the aged asphalt on the rutting factor of the reclaimed asphalt, the dimensionless rutting factor Δ
G*/sinδ was evaluated, and the calculation can be found in Equation (4):
where Δ
G*/sinδ is the dimensionless rutting factor of the reclaimed asphalt, G*/sinδ
mix.r is the relative rutting factor of the reclaimed asphalt, G*/sinδ
mix.r = (G*/sinδ
mix)/(G*/sinδ
new), G*/sinδ
mix is the rutting factor of the reclaimed asphalt, G*/sinδ
new is the rutting factor of the virgin asphalt, G*/sinδ
new.r is the relative rutting factor of the virgin asphalt, G*/sinδ
new.r = 1, and
x is the amount of aged asphalt mixing.
The variance results for Δ
G*/sinδ and G*/sinδ for reclaimed asphalt at different test temperatures and under different aging conditions are shown in
Table 5 with the significance level of
α = 0.05.
As can be seen from
Table 5, the statistical probability
p-value of G*/sinδ is only less than 0.05 under one test temperature, indicating that using G*/sinδ as an indicator to evaluate the high-temperature stability of the reclaimed asphalt under different aging conditions is unable to distinguish the difference between the virgin asphalt and aged asphalt. The four
p-values of Δ
G*/sinδ are less than 0.05, indicating that using Δ
G*/sinδ as an indicator to evaluate the high-temperature performance of the reclaimed asphalt under different temperature and aging conditions can distinguish the differences between virgin asphalt and aged asphalt. Therefore, it is more reasonable to use Δ
G*/sinδ as an indicator.
The variation in Δ
G*/sinδ versus temperature for the reclaimed asphalt is shown in
Figure 6. It can be found that the Δ
G*/sinδ of the reclaimed asphalt decreases gradually with the increase in temperature. The Δ
G*/sinδ values for A30
W, B30
W, and C30
W at the test temperature of 58 °C were 16.0, 11.8, and 9.2, respectively, indicating that the relative rutting factors of the aged asphalt increased by 16.0, 11.8, and 9.2 for each 1% increase in the admixture of aged asphalt. Under the same temperature conditions, the relative rutting factor of A15
W was 14.9, which was not the same as that of A30
W, indicating that every 1% increase in the admixture of aged asphalt produced inconsistent changes in the high-temperature performance of the reclaimed asphalt, even if the aged asphalt was from the same source. Different reclaimed asphalts at different test temperatures and different aging conditions will produce similar test results as above, which are similar to the relative viscosity test results.
The results of the regression analysis show that Δ
G*/sinδ has a good linear relationship with the test temperature (T). The regression equations are all Δ
G*/sinδ = a
T + b (a and b are fitting parameters), as shown in Equations (5) and (6):
From Equations (5) and (6), it can be seen that the ΔG*/sinδ of the reclaimed asphalt is always linearly related to the test temperature, independent of the source and admixture of aged asphalt, and will not change after short-term aging.
3.4. Asphalt Fatigue Factor Evaluation Test Results and Analysis
An increase in rutting factor enhances the ability of the asphalt to resist permanent deformation under high-temperature conditions; however, a high rutting factor can lead to the asphalt being susceptible to cracking under low- and medium-temperature conditions. Therefore, the fatigue factor G*sinδ was introduced to characterize the ability of the asphalt to resist fatigue cracking under medium-temperature conditions after long-term aging. The variation curves of G*sinδ values with temperature for virgin and reclaimed asphalts after long-term aging are shown in
Figure 7.
As can be seen from
Figure 7, the G*sinδ gradually decreases as the temperature increases, and the G*sinδ of the reclaimed asphalt is larger than that of the virgin asphalt. The higher the amount of aged asphalt admixture, the larger the G*sinδ. The results of the nonlinear regression analysis showed that the G*sinδ was exponentially related to the temperature, and the correlation coefficients were all above 0.95.
Here, ΔG*sinδ was selected as the dimensionless fatigue factor indicator for the reclaimed asphalt, and the effect of the relative fatigue factor of the reclaimed asphalt on its resistance to fatigue cracking under medium-temperature conditions was evaluated.
The variance results for the reclaimed asphalt Δ
G*sinδ and G*sinδ are shown in
Table 6.
As can be seen from
Table 6, the effect of the aged asphalt admixture on the fatigue resistance of reclaimed asphalt cannot be evaluated using G*sinδ as an indicator. The three
p-values of Δ
G*sinδ are less than 0.05, indicating that the effects of the source and admixture on the rheological properties of the reclaimed asphalt can be assessed using Δ
G*sinδ as an indicator. Therefore, Δ
G*sinδ is suitable for characterizing the mid-temperature rheology of aged asphalt. The variation in Δ
G*sinδ of the reclaimed asphalt versus temperature is shown in
Figure 8.
It can be seen that the ΔG*sinδ of the reclaimed asphalt increases gradually with the increase in temperature. Even if the source of the aged asphalt is the same, the fatigue resistance of the reclaimed asphalt will vary with every 1% increase in admixture. The different reclaimed asphalts at different temperatures produced similar test results, as described above, which were similar to the relative viscosity and relative rutting factor test results.
The regression analysis showed that the linear relationship between the Δ
G*sinδ and temperature T is correlated well, and their regression equations are both Δ
G*sinδ = a
T + b (a and b are fitting parameters), the calculation equation for which is shown in (7):
From Equation (7), it can be seen that the effects of the source and admixture of the aged asphalt on the fatigue resistance of the reclaimed asphalt can be similarly expressed by the slope and intercept in the linear relationship equation.