**1. Introduction**

Asphalt binder is a material widely used in paving engineering to hold aggregates. It is a special hydrocarbon composite with complex mechanical properties and plays a major role in determining the performance of asphalt pavements. However, asphalt materials tend to oxidize under the influence of environmental factors such as temperature, humidity and solar radiation, leading to changes in its mechanical and rheological properties.

Laboratory and field studies have shown that after oxidization, the stiffness of the asphalt binder increased [1–3]. Previous research using recovered binder from field cores found that high-temperature performance grade (PG) increased 2.4 to 26.6 ◦C for pavements aged between 10 and 82 months [4]. With ageing, the weak attractions of the asphaltene are destroyed, and asphalt molecules change their orientation and become more tightly packed [5]. In addition, the ageing process resulted in stronger associations between asphalt components due to the generation of polar carbonyl groups, which increased the asphaltene fraction [2]. Changes in both its elastic modulus and its viscosity lead to stiffening of the binder.

**Citation:** Wang, H.; Zhu, Y.; Zhang, W.; Shen, S.; Wu, S.; Mohammad, L.N.; She, X. Effects of Field Aging on Material Properties and Rutting Performance of Asphalt Pavement. *Materials* **2023**, *16*, 225. https:// doi.org/10.3390/ma16010225

Academic Editor: Krzysztof Schabowicz

Received: 6 December 2022 Revised: 18 December 2022 Accepted: 20 December 2022 Published: 26 December 2022

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

For the asphalt mixture, numerous laboratory studies have been implemented to investigate property changes with the aging of the asphalt binder. They have shown that the resilient modulus of long-term oven-aged specimens is 50–100% higher than that of short-term oven-aged specimens [2,6]. Moreover, some researchers assumed that due to diffusion resistance, binder oxidation rates in mixtures are slower than that in thin-film neat binders due to diffusion resistance [7–10]. The Hirsch model is often used to describe the relationship between asphalt and the asphalt mixture; however, it was found that due to binder hardening, the mixture ageing actually stiffened the mixture more than the model predicted.

However, few field tests have been performed due to the impact of multiple factors such as climate and binder modification. The current research only focuses on the influence of the dynamic modulus. For instance, the dynamic modulus of field cores in Texas increased by approximately 70% and 130% at the end of 8 and 14 months compared with that at 1 month [3]. On the other hand, such increases are much less, with 27–31% after one year of service in the cold regions of Sweden [11]. Furthermore, it was found that polymer- and rubber-modified binders may reduce the ageing rate in contrast to unmodified binders [4,12,13], and the Evotherm WMA binder generally ages faster than the HMA and Foaming WMA binders [1,3,4].

In summary, most aging studies are focused on laboratory aging, and very few studies have attempted to examine the effects of field aging on the property evolution of asphalt and asphalt mixture. In addition, the field studies in most cases were based on one specific project; thus, the effect of factors such as climate condition, HMA layer thickness and binder types cannot be well analyzed [14,15]. More importantly, limited research exists on how the change in field binder properties over time has affected field mixture properties [16,17]. This study evaluates and quantifies the effect of field aging on the rutting performance of asphalt binders and asphalt mixture over time using field projects from different climatic zones. The study also analyzes the effect of asphalt binder property change on mixture alternation.

#### **2. Methodology**

Four projects selected from the NCHRP 9–49A project report 843 were used for performing the analysis introduced in this paper. These projects were selected to cover different pavement structures and material properties. Within each project, field cores were taken and material properties that are typically used to describe asphalt stiffness and rutting resistance were tested, consisting of binder high temperature PG, binder multiple stress creep recovery (MSCR), mixture dynamic modulus and mixture Hamburg rut depth. These material properties were tested based on field cores taken at two different periods. In the first series, cores were obtained shortly after pavement construction and in the second series cores were obtained two to three years after the construction was finished. All the field cores were taken inside the wheel path to study purely the effect of aging to exclude the effects of traffic load. Other factors that may affect field aging and field rut depth were also collected, consisting of in-place air voids, pavement structure and climate. Additionally, aggregate gradation and asphalt content were obtained to confirm that no significant changes in the cores happened due to traffic. The material properties between the two series of field aging were compared. The effect of asphalt property change on the asphalt mixture properties was also analyzed.

#### **3. Project Information**

Four projects that were constructed in 2011 and 2012 were investigated, which covered different climatic zones, pavement types (HMA and WMA), traffic levels and pavement structures. The four pavement projects were located in four states: Montana, Tennessee, Iowa and Louisiana, and are therefore referred to the MT I-15, TN SR 125, IA US 34 and LA US 61. Three 61-m test sections of HMA and WMA pavements were selected for further study. Field construction information of the four projects that was collected includes:


The key information for the four field projects is concluded in Table 1. Each project includes at least one WMA technology and a homological control HMA pavement. More detailed information of the four projects can be found in the NCHPR report 843 [18].


**Table 1.** Specimen Grading for Indoor Test.

Note: H—HMA, W—WMA, S—Sasobit, E—Evotherm, F—Foaming, NMAS—nominal maximum aggregate size, AADTT—average annual daily truck traffic.

#### *3.1. Pavement Structure*

The pavement structure that was used for each project is shown in Figure 1. The thickness of each pavement layer is presented, and the subgrade soil is assumed to be infinite in depth. Two existing pavement structure types were selected, including flexible pavements and a combination of asphalt/PCC pavements.

**Figure 1.** Pavement structure diagram of each project. (From top to bottom in turn are overlay, existing HMA, base, and subgrade respectively).

#### *3.2. Field Climate Information*

Climate information, consisting of high temperature hour, shortwave solar radiation and humidity are highly related to asphalt aging [2,18,19] and were obtained from the long-term pavement performance (LTPP) website, InfoPave. The number of air temperature hours referred to in this paper is the hours when the pavement temperatures are higher than 25 ◦C, as recommended [19]. As shown in Figure 2, the MT I-15 project has the lowest values of relative humidity, air temperature hour > 25 ◦C, shortwave solar radiation and indicating the least ageing impact. The other three projects experienced similar relative humidity and shortwave solar radiation. The LA US 61 project had the highest value of air temperature hour > 25 ◦C and may have encountered the most severe ageing. Note that the climate information comprises the accumulated values which cover the period of being open to traffic until the second round of field core samples were taken.

**Figure 2.** Field climate information: (**a**) relative humidity; (**b**) air temperature hour > 25 ◦C; (**c**) shortwave solar radiation.

#### **4. Data Collection**

#### *4.1. Field Climate Information*

Cores were taken in the field from the non-wheel path to limit the potential of preexisting damage in the material. In the laboratory, the existing pavement was removed for further core tests. After the core fracture test, binder extraction and recovery were performed using the entire overlay specimen to evaluate the average aging effect without taking the aging gradient variation through overlay thickness into account.

#### *4.2. Binder Extraction and Recovery*

Asphalt binders were extracted based on AASHTO T164 and recovered according to AASHTO R59. The used chemical was a combination of 85% Toluene and 15% Ethanol by volume. Both WMA and HMA on-site cores were heated 110 ◦C until they were loose enough to separate. The separated mixtures were cooled down at room temperature before extraction. The minimum mass of samples used for binder extraction was determined by nominal maximum aggregate size (NMAS). Usually, several extractions were needed until the extract was no darker than a light straw color. Recovered binders were tested by taking them as short-term aged (rolling thin-film oven-aged) asphalt, as suggested [14,15].

### *4.3. Aggregate Gradation and Asphalt Content*

Aggregate gradation and asphalt content were checked using the field cores after fracture tests for both the first-round and the second-round surveys. The asphalt content was determined in accordance with AASHTO T 308, and aggregated gradation was checked following AASHTO T 30.

#### *4.4. Material Properties*

In the laboratory, an overlay of field cores and recovered asphalt binders was used to conduct a series of laboratory tests to determine the physical and engineering material properties. Table 2 shows the summary of all the laboratory mixture and binder tests. The MSCR test temperatures were determined based on the high pavement temperature of specific project locations obtained from LTPPBind Version 3.1 software (Federal Highway Administration, Washington, DC, USA) at 98% reliability.

**Table 2.** Summary of Laboratory Mixture and Binder Testing.


Note: IDT—indirect tension test, HWT—Hamburg wheel track, DSR—dynamic shear rheometer, AMPT—asphalt mixture performance tester, PG—performance grade, MSCR—multiple stress recovery, Jnr—non-recoverable creep compliance, R—percent of recovery, AV—air voids, AC—asphalt content.

#### **5. Results and Analysis**

#### *5.1. Air Voids*

Air voids are defined as the pockets of air between the asphalt-coated aggregate particles in a compacted asphalt paving mixture. Asphalt mixtures of high air voids generally show a faster rate of asphalt aging than the mixture with low air voids [2], which may be due to the fact that asphalt oxidation is the chemical reaction of asphalt with oxygen [5], and high air voids increased the contact areas between air and asphalt mixture.

Figure 3 summarizes the average air voids content based on three core replicates. Error bars that represent standard deviation are also shown. As it can be seen, the air voids from the first round exhibited similar values in general to the second round for most projects (except for MT I-15). This is within our expectations since all the cores were taken at the non-wheel path (with limited traffic load) and the differences in the air voids between the two rounds could be caused by construction variation. For the MT I-15 project, chip seal was placed one year after being open to traffic, which could affect the air voids from the second round.

**Figure 3.** Comparison of in-place air voids.

#### *5.2. Aggregate Gradation and Asphalt Content*

Aggregate gradation is expressed by percent passing of coarse aggregate (aggregate predominately retained on 4.75 mm sieve) and fine aggregate (aggregate almost entirely passing the 4.75 mm sieve). Aggregate gradation helps determining important asphalt pavement properties such as stiffness and rutting resistance. The asphalt content is the ratio between the asphalt weight and the total mixture weight (asphalt plus aggregate). Higher asphalt content generates thicker film thickness and reduces aging effect, whereas lower asphalt content in general increase mixture stiffness and result in better rutting resistance [2].

Table 3 summarizes aggregate gradation determined based on field cores. Note that the aggregate gradation shows a range because each project concludes two or more HMA and WMA pavements; this range covers all the pavements from specific project. As observed, there is no significant gradation difference between the two rounds for all the projects except for the IA US 34. A large percentage of the aggregate in the IA US 34 project is limestone, which has a lower strength and is weaker than the other paving aggregate types such as granite and gravel.

**Table 3.** Aggregate Gradation Percent Passing Comparison Between the First- and the Second-Round Survey.


Asphalt content was also determined based on field cores. The asphalt content difference between the two rounds is smaller than 0.1%; such a small difference should not significantly affect the rutting resistance of asphalt pavement.

#### *5.3. Recovered Binder High-Temperature PG*

The high-temperature PG evaluates the rheology properties of the binder at various temperatures. Typically, a greater high-temperature PG value indicates a stiffer asphalt binder (more rutting resistance) caused by aging or asphalt modification.

Figure 4 illustrates the high-temperature PG comparisons between the first and the second rounds of the extracted binders. For projects examining TN SR 125, IA US 34, and LA US 61, as shown, the high-temperature PG of the second round of extracted binders is always higher than those of the first round, which implies a clear effect of field aging during the two to three years of service.

For the MT I-15 project, it was observed that the PGs of the extracted binders in the second round were close to those in the first round. The cold local climate as shown in Figure 3, plus the chip seals that covered the surfaces, could have reduced the aging effect of the asphalt. The binders of the TN SR 125 project show the highest PG increase over time, which may be explained by the thin asphalt layer applied. The IA US 34 binders have a relatively small high-temperature PG, which indicates less rutting resistance ability.

**Figure 4.** Comparisons of extracted binder high-temperature PGs.

For the LA US 61 project, the high-temperature PG increase is relatively small, considering that the high temperature hour > 25 ◦C is the highest among the four projects. Such slow ageing could be caused by the PG 76-XX polymer-modified asphalt (PMA) binder used. Based on on-site cores from more than 20 field projects, Zhang et al. [4] found that the PG76-XX PMA binders generally aged less than PG64-XX and PG70-XX binders. The reason for less ageing may be that polymers can prevent the formation of sulfoxides on ageing [20].

#### *5.4. Recovered Binder MSCR Test*

The MSCR test is used to evaluate the asphalt binder's potential for permanent deformation. The test is performed by dynamic shear rheometer (DSR) under shear creep and recovery at two stress levels (0.1 and 3.2 kPa) at a specified temperature. The creep portion of the test lasts for 1 s at a constant stress, which is followed bya9srecovery. Ten creep and recovery cycles were tested at each stress level. Two parameters, non-recoverable creep compliance (Jnr) and percent recovery (R), were obtained from the test and can be calculated using:

$$\in\_I(\text{3.2}, N) = \frac{\left(\in\_1 - \in\_{10}\right) \times 100}{\in\_1} \tag{1}$$

$$R\_{3.2} = \frac{SIM[\in\_r \text{ (3.2, N)}]}{10} \text{ for N} = 1 \text{ to } 10 \tag{2}$$

$$J\_{nr}(3.2, N) = \frac{\in\_{10}}{3.2} \tag{3}$$

$$J\_{nr\_{32}} = \frac{SIM[I\_{nr}(3.2, N)]}{10} \text{ for N} = 1 \text{ to } 10 \tag{4}$$

where ∈<sup>1</sup> is the strain value at the end of the creep portion (i.e., after 1 s) of each cycle, and ∈<sup>10</sup> is the strain value at the end of the recovery portion (i.e., after 10 s) of each cycle.

This paper aims at 3.2 kPa stress level since high stress levels are more important for rut depth development. ∈*<sup>r</sup>* (3.2, N) and *Jnr* (3.2, N) indicate percent recovery and nonrecoverable creep compliance at each cycle, respectively. Finally, average percent recovery (*R3.2*) and average nonrecoverable creep compliance (*Jnr3.2*) at 3.2 kPa were obtained by calculating the mean percent recovery and mean nonrecoverable creep compliance from cycles 1 to 10.

Figure 5a,b compare the *Jnr3.2* and *R3.2* between the first- and the second-roundextracted binders, respectively. For the MT I-15 project, no obvious changes in *Jnr3.2* and *R3.2* were observed between the two rounds, which again could be ascribed to the reduced

aging effect due to the chip seal surface treatment. For the other projects, both HMA and WMA binders from the second-round cores show lower *Jnr3.2* and higher *R3.2* values than those in the first round, indicating an improved rutting resistance of the pavements.

**Figure 5.** Comparisons of recovered binder MSCR results (**a**) *Jnr3.2*, and (**b**) *R3.2*. Note: the number shown in the parenthesis indicates test temperatures.

Among the four projects, the IA US 34 project shows the highest *Jnr3.2* and the lowest *R3.2*, which implies the high potential of rutting susceptibility. Asphalt binders from the MT I-15 project show the lowest *Jnr3.2* and the highest *R3.2* values, illustrating good rutting resistance. The *Jnr3.2* and *R3.2* values between the TN SR 125 and the LA US 61 are similar to each other.
