3.1. Determination of the Optimum Bitumen Content (OBC)
The OBC of the mixtures was determined using the Marshall mix design method. According to VicRoads-RC500.01 [29
], the asphalt mix designed for light to medium traffic roads, i.e., type N for the wearing course and type SI for the structural course, must have 4.9 to 5.3% air voids. Thus, the bitumen content corresponding to an air void of 5.0% was selected as the OBC of the mixture and other properties such as the voids in mineral aggregate (VMA), stability, flow, and voids filled with the binder (VFB) were checked against the recommended limits stipulated in VicRoads-RC500.01 [29
] and Asphalt-Institute [30
]. For this, the above-mentioned properties were plotted against the bitumen content. Example Marshall charts developed for the 10NA mixture are shown in Figure 7
Using the Marshall method plots, the VMA, VFB, stability, flow value, and unit weight of the samples at the OBC of each mixture were calculated as demonstrated in Figure 8
and Figure 9
for the 10 and 20 mm mixtures, respectively. As can be seen in Figure 8
and Figure 9
, 10NA, 10RCA 75, 10RCA 65, and 10RCA 55 had 5.2, 6.9, 6.4, and 6.4% OBC respectively. Similarly, 20NA, 20RCA 75, 20RCA 65, and 20RCA 55 had 5.8, 6.6, 6.4, and 6.1% OBC respectively. The OBC results revealed that NA asphalt mixtures have less OBC compared to recycled material mixtures.
The voids in mineral aggregate (VMA) refer to the void space between the aggregate particles in a compacted specimen. Bruce A. Chadbourn [40
] reported that an increase in finer particles decreased the VMA in an asphalt mixture. In this research, the mixtures shifted towards a finer gradation with a simultaneous reduction in RCA and addition of RAP, resulting in the reduction of VMA. Following the observed trends in Part f of Figure 8
and Figure 9
, it could be predicted that the simultaneous reduction in RCA and addition of RAP beyond 55% and 35% would eventually reach a point where the VMA requirements would not be met.
The water absorption test result indicated that the absorbing capacity of coarse RCA (>4.75 mm) is lower than fine RCA (<4.75 mm). Additionally, the bitumen required to coat coarse particles is less than fine particles due to the lower specific surface area of the coarser blends of aggregates. Visual inspection of the aggregates showed that the 10 mm RCA has more roundly shaped particles than the 20 mm RCA. Furthermore, the flakiness index test result shows that the flakiness index of the 10 mm and 20 mm RCA are 15% and 30%, respectively. Flaky particles in an asphalt mix could result in a more compacted mix, possibly due to breakage under compaction, reducing the bitumen consumption. These particles, however, have lower stiffness, which can compromise the mechanical strength of the mixture. The above-mentioned factors result in lower OBCs for the 20 mm recycled material mixtures than those for 10 mm mixtures, while the stability test result remains quite similar.
The stability of an asphalt mix is directly proportional to the friction between particles and cohesion between bitumen and particles. Inter-particle friction depends on the surface roughness, inter-granular contact pressure, bitumen content and grade, aggregate gradation, and angularity, whereas cohesion depends on the aggregate gradation and density [39
]. Figure 8
and Figure 9
also show that the recycled mixtures’ Marshall stability is greater than that of NA mixtures. The higher stability value indicates the mixture has a higher resistance to distortion, displacement, rutting, and shearing stresses. Figure 8
and Figure 9
show no significant difference in stability test results between the 10 mm and 20 mm recycled material mixes; this could be attributed to the higher flakiness index of 20 mm mixtures compared to the 10 mm mixtures.
The flow value of a test specimen is an indication of the maximum vertical displacement reached during the loading up to the peak load. The flow value is recommended to be between 8 and 16 [30
]. All asphalt mixtures had flow values within the limit specified. In the 10 mm mixtures, the flow value was the lowest for NA, followed by RCA 75, RCA 65, and RCA 55. In the 20 mm mixtures, again, the flow value was the lowest for NA, followed by RCA 75, RCA 55, and RCA 65. ASTM_D6927-15 [41
] states that when the flow value is above the upper limit, the mixture is considered too plastic or unstable, and when the flow value is below the lower limit, it is considered too brittle. Hence, based on the obtained flow values, recycled material mixtures are expected to be less brittle, but more plastic than the control mixtures, yet all are within the acceptable range.
VFB represents the percentage of the voids filled with an effective binder in an asphalt mixture. It has an acceptable limit range between 65 to 78% for both 10 and 20 mm asphalt mixtures, respectively [30
]. According to AGPT04B-14 [4
], at a lower VFB, around 60%, the mixture becomes “dry”, lacks cohesion, and exhibits lower durability and fatigue resistance, while at a higher VFB, around 85% or more, the mixture can become unstable and susceptible to rutting. Figure 8
and Figure 9
show that for both sizes of asphalt mixtures, the NA asphalt mixtures had a higher VFB than the recycled material asphalt mixtures. Similarly, 20RCA 65 had the lowest VFB, followed by 20RCA55, 10RCA 55, 20RCA 75, 20NA, 10RCA 75, 10RCA 65, and 10NA with a VFB of 78%. Therefore, based on the VFB test result, the recycled material mixtures are predicted to be less susceptible to rutting.
The bulk density of NA mixtures was higher than recycled material mixtures. This could be attributed to the lower particle density of the recycled materials compared to the natural aggregates used in this research (Figure 3
). With decreased RCA content and increased RAP content in the asphalt mixture, the bulk density of the asphalt mixture at their respective OBC is observed to be increasing.
3.2. Indirect Tensile Modulus (IDT) Test
The IDT test was performed at temperatures 21, 25, and 29 °C. Figure 10
illustrates the resilient moduli of all mixtures at these temperatures. While the resilient modulus generally decreases with the increase in the temperature, the trends demonstrated in Figure 10
for all mixtures except 20RCA 55, 10RCA 65, and 10NA generally reveal a more significant decrease in resilient modulus from 25 to 29 °C than from 21 to 25 °C.
The IDT test result showing the resilient modulus of all asphalt mixtures at 25 °C (as required by AS_2891.13.1 [36
]) with obtained ranges are presented in Table 5
. The results presented in Table 5
reveal that the recycled aggregate mixtures are generally stiffer than the conventional HMA mixtures. The stiffness of the 10 mm HMA increased with the simultaneous reduction of RCA and increase in RAP in the mixture. This could be because of the shift in gradation, which gets finer with the reduction of RCA in the mixture. However, the resilient modulus of 20 mm HMA first increased and then decreased with the reduction in RCA. The 20RCA 65 specimen had a coarser gradation, higher OBC, and lower particle density compared to 20RCA 55. These factors could have increased the effective bitumen content in the mixture, resulting in a stiffer mixture [41
]. The lower IDT of 20RCA 75 could be attributed to the higher absorption potential due to the higher percentage of RCA in the mixture, thereby lowering the effective bitumen content due to the presence of more voids and highly absorptive RCA in a higher proportion.
Al-Mosawe et al. [42
] compared the resilient modulus of thirteen mixtures of different gradation using similar aggregates and concluded that, generally, the denser mixtures tend to have lower air voids and exhibit a higher stiffness. Generally, with the rise in the density of mixtures and reduction in the number of voids, a higher stiffness is expected to be achieved. However, due to the use of three different recycled aggregates in this research, variation of the RCA content in the mixture changes not only the density but also the proportion of angular aggregates in the mixture, as well as the absorption potential and OBC of the mixture. Hence, the mixture density alone could not be relied on to anticipate the increase or decrease in stiffness.
3.4. Moisture Sensitivity
demonstrates the TSR of all asphalt mixtures compacted using the Marshall compactor. The results presented in Table 7
show that all mixtures, except for the two NA mixtures and 10RCA 65, generally met the required limit specified by VicRoads requirements [29
]. Although a TSR value of 80% or greater is considered acceptable, some agencies have chosen to accept TSR values of 70% or greater based on their experience [30
]. It should be noted that the main goal of this experiment was comparing the performance of control mixtures (NA specimens) with that of recycled material mixtures prepared and tested in the same environment using the same procedures. Thus, although NA mixtures should ideally meet the requirements, the obtained values can still be used for comparison purposes. The results also show that recycled material mixtures generally have greater resistance to moisture damage than natural aggregate mixtures. This could partially be due to the presence of unhydrated cement in the RCA that contributes to the binding of the mixture over time when submerged in water.
It can also be observed from the results of Table 7
that, in RCA75 and RCA65 mixtures, the 20 mm samples exhibit greater TSR. In conventional asphalt mixtures made of virgin aggregates, TSR is generally known to be higher for mixtures with smaller maximum particle sizes [43
]. This is supported by the TSR results of the NA mixtures of the current study. The mixtures of this study on the other hand, are made of three different materials with different shapes, surface textures and maximum particle sizes: RCA (maximum size = 20 mm), RAP (maximum size = 10 mm) and RG (maximum size = 4.74 mm). In the 10 mm mixtures, coarse particles of RCA (>10 mm) which have a relatively rough surface texture (in contrast to the RAP particles and RCA particles < 10 mm) are excluded. This could be the reason for the higher TSR of 20 mm RCA 75 and RCA65 mixtures, as a rougher particle surface can result in a stronger bond when bitumen is added to the mix. However, by increasing the RAP content (and hence reducing the RCA content) in RCA55 the proportion of the rough >10 mm RCA particles is reduced and thus the exclusion of this small proportion in the 10 mm mixtures does not result in a significant difference between the TSR values. Furthermore, the TSR values in recycled material mixtures do not follow a specific trend in terms of an increased or decreased RCA content. This could be due to the inhomogeneity of RCA particles as also reported by Bastidas-Martínez et al. [17
]. It is expected that by improved recycling technologies and methods in future, more quality-consistent recycled materials are produced, to match the proportion of current virgin aggregates.
demonstrates that all asphalt mixes except for 10NA and 10RCA75 satisfy the minimum wet tensile strength criteria specified in VicRoads technical note [23
]. With the decrease in RCA content in an asphalt mix, the minimum wet tensile strength of an asphalt mix was observed to be increasing. However, the minimum wet tensile strength of 20RCA 65 is slightly higher than 20RCA 55.
Comparing the recycled and natural aggregate asphalt mixtures in terms of the tensile strength ratio, 10RCA 75 and 10RCA 55 performed better than 10NA. In the case of 20 mm asphalt mixtures, all three recycled material aggregate asphalt mixtures exhibited higher TSR than the 20NA mixture. In terms of minimum wet tensile strength, all recycled material asphalt mixtures showed greater wet tensile strength than natural aggregate specimens. The wet tensile strength is of particular importance in areas with a wet climate, such as Melbourne, Australia.