Effect of Reclaimed Asphalt Pavement Heating Temperature on the Compactability of Recycled Hot Mix Asphalt

The compactability of an asphalt mixture is related to the heating temperature of the materials, but the heating temperature of reclaimed asphalt pavement (RAP) is limited by the production process of hot-in-plant recycled mixtures. To choose a reasonable heating temperature for RAP according to the compactability, the compaction energy ratio (CER) obtained from the Superpave gyratory compactor compaction curve was developed. The CERs of fourteen kinds of asphalt mixtures made with different RAPs were compared, all of which were different in type, content, and heating temperature. The results indicated that CER is an effective energy index to evaluate the workability of a bituminous mixture, and it considers both the accumulated energy after each gyration and the number of gyrations. It was also found that increasing the heating temperature of the RAP cannot always improve the workability of the recycled mixture, because the higher heating temperature caused more hard-aged bitumen to be blended with soft virgin bitumen during the mixing process. At the same RAP heating temperature, increasing the RAP content made it more difficult to compact the mixture, especially for RAPs with styrene–butadiene–styrene (SBS) modified bitumen, and the recycled mixtures with SBS-modified bitumen were more difficult to compact than those with nonmodified bitumen.


Introduction
With the increasing demand for asphalt pavement rehabilitation, an increasing amount of reclaimed asphalt pavements (RAPs) are being produced and need to be recycled. As RAP contains valuable asphalt binders and aggregates, they can replace the virgin binder, and can replace aggregate in new asphalt pavement construction and asphalt pavement rehabilitation, providing significant environmental and economic benefits [1]. The recycling techniques are a hot research topic in the field of road engineering [2][3][4]. According to the recycling processes and mixing temperature, the recycling methodologies can be classified as hot recycling (HR) and cold recycling (CR) [5]. Although CR possesses superior advantages in its environmental effect [6], the HR recycling technique has a wide range of applications because of its improved performance [7].
As with fresh hot mix asphalt (HMA), temperature is considered a key factor affecting the compactability of recycled HMA, and poorly compacted mixtures may lead to severe rutting, poor fatigue resistance, short durability, and premature moisture damage [8,9]. Increasing the heating temperature of RAP is typically performed in a drum mix plant so as to increase the temperature of recycled RAP.

Materials
To achieve a more systematic evaluation, two typical RAP materials used in China were selected in this study. The first one is the AC-13, which is a surface layer mixture with a nominal maximum aggregate size (NMAS) of 13.2 mm. It was made with SBS modified bitumen and basalt aggregate. The second one is an AC-20 mixture, which is a middle/lower layer mixture with an NMAS of 19.0 mm. It was made with nonmodified base bitumen and limestone aggregate.
Two types of recycled HMA were produced from the two RAPs, namely: recycled AC-13 and recycled AC-20. The recycled AC-13 was made with AC-13 RAP, SBS modified bitumen, basalt aggregate, and mineral filler, while the recycled AC-20 was made with AC-20 RAP, base bitumen, limestone aggregate, and mineral filler. The recycled HMAs were produced at three RAP percentages (0, 20%, and 40%) and three RAP heating temperatures (100 • C, 120 • C, and 140 • C).
Based on the aforementioned testing variables, 14 HMA mixtures were prepared in a laboratory, as shown in Table 1. The recycled mixture was given a coded name according to the NMAS, percentage of RAP (%), and the RAP heating temperature ( • C), for example, A-B-C, where A represents NMAS, B represents percentage of RAP (%), and C represents RAP heating temperature ( • C). Among the most important properties of asphalt mixtures is the binder content. In order to determine the binder content of RAP, extraction tests were conducted using a centrifuge with trichloroethylene as the solvent, in accordance with Chinese specification T 0722-1933 [19].

Gradation analysis
After the binder extraction tests, the remaining mineral mixture was placed in an oven at 105 • C to dry. After drying, the mineral mixture was cooled to room temperature and sieve analysis was conducted to determine the aggregate gradation of RAP.

Binder recovery and characterization
To evaluate the aged binder in the RAPs, RAP samples were first dissolved in a toluene solvent to remove all solid particles with filter and centrifuge. Then, the RAP binder was recovered by vacuum distillation using a rotary evaporator, according to the Chinese specification T 0727-2011 [19]. Finally, the recovered binder was characterized through penetration (T 0604-2011), ductility (T 0605-2011), and softening point tests (T 0606-2011) [19].

Recycled HMA Mix Design
•

Gradation design
The recycled HMA with the same NMAS and the same target gradation was designed according to the Chinese specification JTG F40-2004 [20]. In the gradation design of recycled HMA, the aged aggregate was seen as a kind of aggregate based on the particle size distribution. We fixed the percentage of aged aggregate to be the same as the percentage of RAP and adjusted the proportion of new aggregate to make the synthetic gradation of the aged and virgin aggregate to be close to the target gradation. •

Optimum binder content
The optimum binder content of the recycled HMA was determined by the Marshall method (75 blows per side), according to the Chinese specification JTGT 5521-2019 [21]. In the process of mix design, the heating temperature of the RAP was 120 • C.

Compactability Evaluation
Superpave gyratory compaction is considered to be one of the best methods to assess the compactability of asphalt mixtures [22][23][24]. For this reason, a gyratory compactor was used to investigate the compaction characteristics of different recycled mixtures. The method of preparation of the specimens used the following procedure: 1.
The RAPs were put into an oven at the needed temperature to achieve the desired heat; the heating time was not more than 2 h so as to avoid the RAP being further aged.

2.
The heating temperature of the virgin aggregate was higher than the mixing temperature of 10-15 • C; the limestone was heated to 175 • C and the basalt was heated to 165 • C. 3.
The order of addition of the materials was firstly a virgin aggregate and RAP mixed in the mixer for 1 min; secondly, the new binder was added; and, finally, the heated filler was added. Then, all of the materials were mixed until homogeneous, and the total mixing time was generally 3 min.

4.
After the mixing process, the specimens were compacted using the Superpave gyratory compacter.
A total of 160 gyrations were applied to all of the mixtures, with a compaction pressure of 600 kPa and a compaction angle of 1.25 • . During the gyratory compaction, the relationship between the number of gyrations and the height of the specimen was recorded, which is referred to as the compaction curve. Based on the compaction curve, the bulk volume density of the specimen after each gyration could be estimated from its height. Then, the degree of compaction (DC) could be calculated, which was equal to the percentage of bulk volume density to the theoretical maximum density of the compacted mixture. The air void content (V) and DC held the following relation: where V i is the air void content for a given number of gyration (%), and DC i is the degree of compaction for a given number of gyration (%).

Compaction energy index
In the literature, many of the indices obtained from the SGC compaction curves have been reported and used to evaluate the asphalt mixture compactability, and the compaction energy index (CEI) and traffic densification index (TDI) were used generally [23,25], which were determined using the compaction curve data (as illustrated in Figure 1) for measuring compaction energy [26,27]. The CEI index focused on the results from cycle 8 until a DC of 92% was reached, which corresponded to the minimum density for traffic opening [28]. The TDI index was related to the possible postdensification as a result of the traffic effects. It was determined from the data within the DC range of 92% to 98% [29]. Within this range, the mixture approached the plastic behavior zone.
The CEI value indicates the ease of laying mixes during construction, which is used to evaluate the compactability of HMA. The lower this value, the more easily the mixture can be compacted.
As indicated, the relation between the accumulated compaction energy and the number of gyrations follows a simple linear relation. As a result, Elsa [23] proposed a linear equation to fit the data, namely, Equation (2): where A A is the accumulated area, and a and b are the regression coefficients.  The CEI value indicates the ease of laying mixes during construction, which is used to evaluate the compactability of HMA. The lower this value, the more easily the mixture can be compacted.
As indicated, the relation between the accumulated compaction energy and the number of gyrations follows a simple linear relation. As a result, Elsa [23] proposed a linear equation to fit the data, namely, Equation (2): where AA is the accumulated area, and a and b are the regression coefficients.

 Compaction energy ratio
It is worth noting that the CEI cannot directly be used to determine the change of compaction energy in the process, and the number of gyrations to reach a DC of 92% has a significant effect on the value of CEI. For example, in Figure 2, the values of CEI are equal for the two curves, but it is obvious that the mixture represented by the lower curve has a better compactability, because its DC was higher after the same number of gyrations at this stage. In this study, the energy index was modified to consider the effect of each cycle of gyration on the compaction characteristics [30]. As illustrated in Figure 3, the area of the adjacent numbers of

•
Compaction energy ratio It is worth noting that the CEI cannot directly be used to determine the change of compaction energy in the process, and the number of gyrations to reach a DC of 92% has a significant effect on the value of CEI. For example, in Figure 2, the values of CEI are equal for the two curves, but it is obvious that the mixture represented by the lower curve has a better compactability, because its DC was higher after the same number of gyrations at this stage. The CEI value indicates the ease of laying mixes during construction, which is used to evaluate the compactability of HMA. The lower this value, the more easily the mixture can be compacted.
As indicated, the relation between the accumulated compaction energy and the number of gyrations follows a simple linear relation. As a result, Elsa [23] proposed a linear equation to fit the data, namely, Equation (2): where AA is the accumulated area, and a and b are the regression coefficients.

 Compaction energy ratio
It is worth noting that the CEI cannot directly be used to determine the change of compaction energy in the process, and the number of gyrations to reach a DC of 92% has a significant effect on the value of CEI. For example, in Figure 2, the values of CEI are equal for the two curves, but it is obvious that the mixture represented by the lower curve has a better compactability, because its DC was higher after the same number of gyrations at this stage. In this study, the energy index was modified to consider the effect of each cycle of gyration on the compaction characteristics [30]. As illustrated in Figure 3, the area of the adjacent numbers of In this study, the energy index was modified to consider the effect of each cycle of gyration on the compaction characteristics [30]. As illustrated in Figure 3, the area of the adjacent numbers of gyration, Ai, could be calculated. This area reflects the change of DC after one gyration cycle. The larger the area, the greater the change of DC, and the HMA is more easily compactable. gyration, Ai, could be calculated. This area reflects the change of DC after one gyration cycle. The larger the area, the greater the change of DC, and the HMA is more easily compactable. Under this curve, the accumulated area at N cycles of gyration ( ) can be calculated as follows: (3) Figure 4 shows the relation between the accumulated area (AA) and number of gyrations. A regression equation can be obtained through a logarithmic fitting using Equation (4).
where N is the number of gyrations, is the accumulated area after N cycles of gyration, and a and b are the regression coefficients. As the area at the given adjacent number of gyrations can only reflect the compaction characteristics at this point, the accumulated area and the number at this stage should be considered. Therefore, in the study, the compaction energy ratio (CER) was developed to evaluate the compaction Under this curve, the accumulated area at N cycles of gyration (A A N ) can be calculated as follows: Figure 4 shows the relation between the accumulated area (A A ) and number of gyrations. A regression equation can be obtained through a logarithmic fitting using Equation (4).
where N is the number of gyrations, A A N is the accumulated area after N cycles of gyration, and a and b are the regression coefficients.
Materials 2020, 13, x FOR PEER REVIEW 6 of 14 Materials 2020, 13, x; doi: FOR PEER REVIEW www.mdpi.com/journal/materials gyration, Ai, could be calculated. This area reflects the change of DC after one gyration cycle. The larger the area, the greater the change of DC, and the HMA is more easily compactable. Under this curve, the accumulated area at N cycles of gyration ( ) can be calculated as follows: (3) Figure 4 shows the relation between the accumulated area (AA) and number of gyrations. A regression equation can be obtained through a logarithmic fitting using Equation (4).
where N is the number of gyrations, is the accumulated area after N cycles of gyration, and a and b are the regression coefficients. As the area at the given adjacent number of gyrations can only reflect the compaction characteristics at this point, the accumulated area and the number at this stage should be considered. Therefore, in the study, the compaction energy ratio (CER) was developed to evaluate the compaction As the area at the given adjacent number of gyrations can only reflect the compaction characteristics at this point, the accumulated area and the number at this stage should be considered. Therefore, in the study, the compaction energy ratio (CER) was developed to evaluate the compaction characteristics of HMA. Assuming A and B represent different number of gyrations, and B is larger than A, the following equation can be used to calculate the CER between A and B (CER A to B ): where A A A is the accumulated area at A cycles, and A A B is the accumulated area at B cycles. When the number of gyrations increases from A to B, CER A to B can reflect the average change of the compaction degree after one cycle, and the value is not affected by the DC at A cycles.

Virgin Material Properties
Tables 2 and 3 present the properties of the virgin binder and the two types of aggregate used in this study. It can be observed in Tables 2 and 3 that the modified bitumen had less penetration than the base bitumen, a higher softening point and higher viscosity, and basalt had a larger density and strength than limestone.

RAP Characterization
As Table 4 shows, the binder contents for AC-13 RAP and AC-20 RAP were 4.7%, and 4.0%, respectively, and the gradations of both RAPs were obviously finer than the recommended gradation of the Chinese specification, which was expected as a result of the breakage of aggregates during the milling process. As Table 5 shows, for both the modified binder and base binder, the recovered bitumen was stiffer than the virgin bitumen (Table 2), as a result of aging.  Figure 5 shows the gradation of both the virgin and recycled HMA. It can be seen that the gradations of the recycled HMA were very close to the virgin HMA. Therefore, the influence of gradation on the compaction characteristics can be weakened. As Table 5 shows, for both the modified binder and base binder, the recovered bitumen was stiffer than the virgin bitumen (Table 2), as a result of aging.  Figure 5 shows the gradation of both the virgin and recycled HMA. It can be seen that the gradations of the recycled HMA were very close to the virgin HMA. Therefore, the influence of gradation on the compaction characteristics can be weakened.  According to JTG F40-2004 [20], the optimum bitumen content (OAC) can be calculated based on the relationship between the bitumen content and the Marshall test indexes, including bulk density, air void, voids in the mineral aggregate (VMA), voids filled with asphalt (VFA), Marshall stability, and Marshall flow. Table 6 presents the Marshall mix design results of all of the HMAs prepared in this study, and OAC is the quality ratio of virgin asphalt to mixture. It can be seen that the actual OAC of the recycled HMA was slightly larger than the predicted asphalt content (P nb ), probably because some aged bitumen was inactive. This phenomenon is in agreement with some previous studies' findings [31][32][33]. Furthermore, the stability of the virgin HMA was lower than those of the recycled HMA, and the difference was more significant at a higher RAP percentage. The Marshall quotient is the ratio of Marshall stability over the flow value; RAP can increase the quotient of HMA, which means the use of RAP can improve the stiffness of the HMA.

Compactability of Recycled HMA
An air void content of 8% is usually regarded as the maximum allowable air void content for the compaction quality control of asphalt pavements. With respect to mixture design, a void content of 4% is usually desired. Therefore, in this study, two sections were considered when analyzing the compactability: one section was the section from N = 8 to V = 8%, and the other section was from N = 8 to V = 4%.
In order to calculate the values of CEI and CER for a given section, the corresponding number of gyrations should be known. As in other studies [24,26], the following equation can be used to explain the relation between air void and the number of gyrations: where V i is the air void content for a given number of cycles (%), V 1 is the air voids content calculated at the first gyration, and K is the compactability factor. N i is the number of gyrations. The corresponding number of gyrations can be calculated using Equation (6), and the regression results are listed in Table 7. The regression coefficients of K and V 1 were used to explain the ease of compaction, but V 1 is mainly related to the initial accumulation state of the mixture, and K can only be used to compare mixes if they have the same V 1 . Therefore, when using the indices of K and V 1 to compare the compactability of different asphalt mixtures, the mixtures should have the same initial accumulation state. As can be seen from Table 7, the virgin mixture did not have the highest K or the lowest V 1 , which contradicts common sense. It might be because the virgin mixture had a different initial accumulation state than the recycled HMA.
For AC-20 HMA, the virgin mixture required more gyration cycles to reach the 8% air void content than the recycled mixture with 20% RAP heated to 140 • C, which contradicts common sense, as described above. The reason is that during the early compaction stage, the contrast relationship between the void content was greatly affected by V 1 , and the virgin mixture had a higher V 1 than the recycled mixture with 20% RAP heated to 140 • C. As the number of gyrations increased, the effect of V 1 decreased, and the contrast in the relationship was greatly affected by the compactability; therefore, it can be seen that the virgin mixture required less gyration cycles to reach the 4% air void content than the recycled mixture with 20% RAP heated to 140 • C, which contradicts common sense, as described above. Therefore, the number of gyrations to reach the target void content is not always a good index to evaluate the compactability of HMA, because it is difficult to determine when the effect of V 1 becomes smaller.
Fitting gyratory compaction data using Equation (2) led to the regression results listed in Table 8. CEI 1 is the accumulated compaction energy from N = 8 to V = 8%, and CEI 2 is the accumulated compaction energy from N = 8 to V = 4%. It is obvious from Equation (2) that the number of gyrations affected the accumulated compaction energy, but the number of gyrations to reach the target void content was affected by V 1 . As can be seen from Table 5, the virgin mixture did not always have the smallest CEI 1 and CEI 2 . For the same NMAS mixture, the CEI 1 and CEI 2 had a different order between the different recycled HMAs. Therefore, CEI is also not a good index to evaluate the compactability of HMA.
CER as an effective and better index was developed to evaluate the compactability of HMA. Fitting gyratory compaction data using Equation (4) led to the regression results listed in Table 9. CER 1 is the compaction energy ratio from N = 8 to V = 8%, and CER 2 is the compaction energy ratio from N = 8 to V = 4%. Author Contributions: Conceptualization, X.M. and P.Z.; methodology, X.M. and Z.L.; validation, X.M. and L.W.; formal analysis, L.W.; investigation, L.W. and P.Z.; data curation, X.M.; writing (original draft preparation), X.M. and P.Z.; writing (review and editing), X.M. and Z.L. All authors have read and agreed to the published version of the manuscript.