*5.5. Field Core Dynamic Modulus*

The dynamic modulus is defined as the complex modulus absolute value calculated by dividing the peak-to-peak stress by peak-to-peak strain for a material of a sinusoidal loading on a material. The dynamic modulus is a performance-related property that can evaluate the mixture and characterize the stiffness of asphalt mixtures for mechanistic–empirical pavement design.

The indirect tension dynamic modulus test was conducted to determine the dynamic modulus while considering the limitations of the core geometry. A sinusoidal compressive loading was applied to the diametric axis of an unconfined cylindrical test specimen. Test temperatures and frequencies are shown in Table 2. The loading was applied on each sample to achieve the target strain levels (40–60 horizontal microstrain and < 100 vertical microstrain) in the linear viscoelastic region [21–23]. The load–deformation mathematical relationship in the indirect tension-loading mode is given by:

$$|E^\*| = \frac{2P\_0}{\pi ad} \frac{\beta\_1 \gamma\_2 - \beta\_2 \gamma\_1}{\gamma\_2 V\_0 - \beta\_2 Ul\_0} \tag{5}$$

where *P*<sup>0</sup> is the peak-to-peak load in N, a indicates loading strip width measured in meters, *d* means the thickness of specimen in meters and *V*<sup>0</sup> and *U*<sup>0</sup> represent peak-to-peak vertical deformation and peak-to-peak horizontal deformation in meters, respectively. *γ*1, *γ*2, *β*1, and *β*<sup>2</sup> are geometric constants.

Figure 6 presents the dynamic modulus test results at a test temperature of 30 ◦C and a test frequency 0.1 Hz since the asphalt pavement is more susceptible to rutting under relative high temperatures and a low frequency [24]. It is observed in the figure that most projects show an obviously higher dynamic modulus of the second-round field cores than that from the first-round. Since all the cores were taken from the non-wheel path, this dynamic modulus increase is presumably due to the significant field aging of the asphalt binder. The MT I-15 project shows the least increase in the dynamic modulus, which could be affected by the chip seals applied. The high amount of asphalt used in chip seal fills voids of overlay and may increase film thickness and asphalt content, thereby reducing field aging and resulting in a similar dynamic modulus between the two rounds.

**Figure 6.** Comparisons of field core dynamic modulus at 30 ◦C and 0.1 Hz.

It is also seen that the field aging increases dynamic modulus up to 800% in the second round for IA US 34 HMA compared with the value from the first round. Meantime, the high temperature PG and the MSCR *Jnr3.2* and *R3.2* of the same pavement changed by 12.8%, 69% and 165%, respectively. This finding denotes that the increase rates of the asphalt binder and the dynamic modulus are not the same, which will be discussed below.

#### *5.6. Field Core HWT Test Results*

The HWT is a widely used test method to determine the rutting resistance and moisture susceptibility of asphalt mixture due to weakness in the aggregate structure, inadequate binder stiffness, or moisture damage. This method measures the rut depth and number of passes to failure and provides information about the rate of permanent deformation from a moving, concentrated load.

The HWT was performed following AASHTO T 324. All tests were conducted at a temperature of 50 ◦C under wet conditions. The speed of the wheel was set as 52 passes per minute. The test terminated when either the rut depth achieved 12.5 mm or a pass number of 20,000 was reached.

Figure 7 summarizes the HWT rut depth at 10,000 passes. This pass number was selected because all the first-round cores from the IA US 34 project reached the test threshold value of 12.5 mm at 10,000 passes. It is observed that in general, the rutting resistance of the second-round core is higher than that from the first round, except for the MT I-15 in which the chip seal may have reduced the aging effect. Since the aggregate gradation and asphalt content between the two rounds did not change greatly, the improved rutting resistance should have been contributed to to a major extent by asphalt aging. Bonding between asphalt and aggregate particles provide significant force in resisting mixture to deform, less aged asphalt is more flow and provides a slip plan between aggregates which facilitates aggregate movement and mixture is easy to deform. In contrast, flow conditions of aged asphalt reduced due to increased viscosity, which helped produce better adhesion between aggregates, and aggregate movement became more restricted. In this case, deformation lessened.

**Figure 7.** Comparisons of field core Hamburg rut depth.

It is also seen that the field aging increases the dynamic modulus up to 800% in the second round for IA US 34 HMA compared with the value from the first round. Meanwhile, the high-temperature PG and the MSCR *Jnr3.2* and *R3.2* of the same pavement changed by 12.8%, 69% and 165%, respectively. This finding denotes that the increase rates of the asphalt binder and the dynamic modulus are not the same, which will be discussed below.

#### *5.7. Effect of Asphalt Aging on Mixture Properties*

In order to study the effects of the variation in asphalt binder properties due to aging on material properties of asphalt mixture, changes to them between the first and second rounds were calculated. These are the increase in high-temperature PG, decrease in MSCR *Jnr3.2*, increase in MSCR *R3.2*, increase in dynamic modulus and decrease in HWT rut depth, respectively.

As seen in Figure 8, the asphalt property changes with dynamic modulus increase correlate well with ageing. The dynamic modulus values increased with the increase in hightemperature PG and MSCR *R3.2*, and the decrease in MSCR *Jnr3.2*. The magnitude changes in the binder properties between the two rounds are 1.2–8.4 ◦C for high-temperature PG, 0.1 to 24.6% for MSCR *R3.2* and 0.02 to 1.07 kPa for MSCR *Jnr3.2*. Those changes corresponded to an increase in dynamic modulus up to 1356 MPa.

**Figure 8.** Effect of asphalt binder property change on dynamic modulus change (**a**) increase in high-temperature PG, (**b**) decrease in MSCR *Jnr3.2* and (**c**) increase in MSCR *R3.2*.

Figure 9 shows the relationship between the binder property change and HWT rut depth. As noted, there was a loose relationship denoting that rut depth in general increased with the increase in high-temperature PG and decrease in MSCR *Jnr3.2*. No correlation between HWT rut depth and MSCR *R3.2* over time was found. As it can be seen, the increase in high-temperature PG (1.2 to 8.4 ◦C) and the decrease in MSCR *Jnr3.2* (0.02 to 1.07 kPa) reduced the HWT rut depth to the maximum value of 7.3 mm between the two rounds.

**Figure 9.** Effect of asphalt binder property change on HWT rut depth change (**a**) increase in hightemperature PG, (**b**) decrease in MSCR *Jnr3.2*, and (**c**) increase in MSCR *R3.2*.

Additionally, the percentage increase/decrease in each material property between the two rounds was calculated and shown in Figure 10. As it can be seen, regarding binder properties, the high-temperature PG has the least percentage increase with the maximum value of 32.2%, whereas the MSCR *R3.2* experienced the most increase with the maximum value of 325%. As for the mixture, the dynamic modulus and the HWT rut depth increased up to 673.1% and 64%, respectively. The large range of MSCR *R3.2* and mixture dynamic modulus make the prediction more complicated. In addition, it is expected that the laboratory ageing could be harder for MSCR *R3.2* due to its large variation.

**Figure 10.** Percentage increase/decrease in each material property.

It is evident that the change in binder property is not proportional to the change in dynamic modulus. Similarly, the change in dynamic modulus is not comparable to the change in the HWT rut depth. Therefore, it is expected that when different binder properties are applied to predict mixture performance, the shift factor would be varied greatly.

#### **6. Conclusions**

This paper evaluated the effect of field pavement ageing on properties of asphalt binder and asphalt mixture at two different rounds. The relationship between property changes in the asphalt and asphalt mixture due to ageing was also analyzed.

Results indicate that asphalt pavement in general became stiffer after years of service in terms of asphalt binder (high temperature PG, MSCR) and asphalt mixture (dynamic modulus, HWT rut depth). The asphalt mixture stiffening is caused to a large degree by asphalt aging, considering that the evaluated cores had no significant differences between the two rounds in in-place air voids, aggregate gradation and asphalt content. Such findings were confirmed by observing a good relationship between the change in binder properties and mixture properties. However, the application of the chip seal significantly reduced the aging process.

The material properties changed to different extents over time. The parameters that are most sensitive to field ageing are MSCR *R3.2* and dynamic modulus. Thus, the comparison among lab-tested material properties may be better checked in rank instead of absolute values. Note that the high variation in MSCR *R3.2* could cause some issues for laboratory aging, which needs further research. The effect of asphalt aging on pavement mixture properties may not follow a proportional liner trend.

The PG76-22 polymer-modified binders in general aged slower than other binder types evaluated in this study, for both asphalt binder and asphalt mixture properties. Since it is already proved that PG 76-22 polymer-modified binders show very good rutting resistance, such slow aging could also be beneficial to cracking resistance.

However, there are also some limitations that need improvement in future studies. For example, only two existing pavement structure types were selected, including flexible pavements, and a combination of asphalt/PCC pavements. For other types of asphalt pavement, the effects of field aging on material properties and rutting performance and their mechanisms need to be further studied.

**Author Contributions:** The authors confirm contribution to the paper as follows: conceptualization, H.W., Y.Z. and W.Z.; data curation, S.S. and S.W.; formal analysis, Y.Z. and L.N.M.; funding acquisition, W.Z.; investigation, S.S. and S.W.; methodology, H.W. and X.S.; writing—original draft, W.Z. and L.N.M.; writing—review and editing, X.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by National Key Research and Development Project with grant number 2020YFB1600102. This study was also sponsored by the National Cooperative Highway Research Program 09-49A. The authors also appreciate the support from the National Natural Science Foundation of China, grant number 52278443.

**Data Availability Statement:** All data, models, or codes that support the findings of this study are available from the corresponding authors upon reasonable request, including the field test data of the asphalt layers.

**Acknowledgments:** The authors would like to acknowledge and thank Ed Harrigan of the NCHRP staff and panel members for their assistance. Thanks also go to Braun Intertec, Inc. and Bloom Companies, LLC, who conducted the field activities, and to partner universities and highway agencies for their generous help.

**Conflicts of Interest:** The authors declare no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
