**1. Introduction**

Construction and demolition (C&D) waste management has become a worldwide concern, as up to 600 million tons of waste construction materials are produced each year. Generally, C&D waste mainly contains coarse and fine sand aggregates, aging asphalt, hardened cement hydrate, and other components [1–4]. C&D waste-recycling research began near the end of the 20th century in the United States, when the Texas Department of Transportation initially investigated the feasibility of using a waste–asphalt mixture in highway construction and maintenance application in 1994. A large number of waste materials were attempted to be used in sub-grade fillers, pavement bases, and other infrastructure construction. The United States saved 4.1 million tons of matrix asphalt and 78 million tons of natural stone in 2018 [5,6]. Similarly, the European Asphalt Pavement Association suggested all its member countries use recycled waste–asphalt materials in 2002. Over 90% of waste–asphalt mixtures have been used for pavement and pavement base materials [7–9]. However, with the development of urbanization, about 1.7 billion tons of C&D waste (2019) was generated in China, and the recycling rate is far lower than that in developed countries [10–13]. Thus, using reclaimed C&D waste in new highway construction is a promising way to solve these imminent issues.

As the largest part of ordinary concrete mixes, the excessive consumption (approximately 26 billion tons per year) of nature aggregate (NA) leads to harmful environmental pollution, and a possible solution to improve the sustainability and cost-effectiveness between C&D and NA needs to be investigated [14–17]. The reuse of recycled concrete aggregate (RCA) and recycled asphalt pavement (RAP) aggregate in highway rehabilitation has attracted extensive attention in recent years. The RAP aggregate is a mixture of aggregate and bitumen, mainly derived from old asphalt pavement. Due to the asphalt

**Citation:** Xu, W.; Li, W.; Ji, Y. Mechanical Behavior Investigation of Reclaimed Asphalt Aggregate Concrete in a Cold Region. *Materials* **2021**, *14*, 4101. https://doi.org/ 10.3390/ma14154101

Academic Editor: Carlos Morón Fernández

Received: 11 June 2021 Accepted: 21 July 2021 Published: 23 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 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/).

binder, the RAP aggregate has a worse environmental effect and weaker inner bond when compared with RCA.

Existing studies focus on RAP's material properties and optimal replacement ratio with NA [18–22]. Saeed and Reza [23] evaluated the performance of recycled asphalt mixtures in C&D waste materials, and the optimal binder content was determined. Test results showed that the rutting resistance was effectively improved by 30% in recycled aggregates. Akash et al. [24] used rheological and chemical methods to investigate asphalt binder and mixtures, and various recycling agents were divided into three categories. A novel parameter was developed to predict the effectiveness of various recycling agents. Wojciech [25] evaluated the fatigue life of ASP in an asphalt–concrete mixture, and several lab tests (air-void content, penetration, stiffness) were made to evaluate the mixture's mechanistic performance. A French method was presented to calculate the mixture's fatigue life. Juntao et al. [26] tested an eight-year recycled asphalt mixture bound by emulsion, and the long-term performance and interface microstructure were discussed. Test results showed that the tensile strengths and creep deformation met the performance requirement. Hassan [27] studied the behavior of reclaimed asphalt pavement (RAP) aggregate concrete, and various mixtures (natural aggregate, reclaimed coarse aggregate, reclaimed coarse and fine aggregate, and reclaimed coarse aggregate with 30% fly ash) were selected to find the optimum performance. Research results showed that the RAP aggregate partially reduced the concrete's mechanical performance (compression and tensile strength). However, the properties of ductility and microstructure were improved due to the effect of fly ash and RAP aggregate. Papakonstantinou [28] investigated the performance of recycled asphalt pavement (RAP) aggregate use in Portland cement concrete (PCC). Five weight percentages (5%, 7.5%, 10%, 12.5%, and 15%) of RAP aggregates were used in concrete mix design, and their mechanical performance was examined. The test results found that the compression strength and elastic modulus had a negative relation with increasing RAP ratio, and all mixtures met the requirements for road performance. Zaumanis et al. [29] proposed a performance-based design method to solve the asphalt mix design procedures, and the key parameters were determined to improve the asphalt mixture's performance.

Other studies investigated RAP concrete applications, including new asphalt pavement, base layers, structural members, and so on [30,31]. Giulia et al. [32] summarized the development of reclaimed asphalt pavement (RAP) material used in new asphalt construction, and the effect of RAP content was discussed. Abdulgazi [33] discussed the potential utilization of construction demolition waste (CDW) using in hot-mix asphalt pavements, and alternative CDW technical specifications and guidelines were presented that mainly considered safety, tolerability, and efficacy. Fawaz et al. [34] summarized the current knowledge about reclaimed asphalt pavement (RAP) and recycled asphalt shingles (RAS) in the United States, and the current and future challenges for reclaimed asphalt utilization were presented. Sharareh et al. [35] evaluated the effects of recycled asphalt shingles (RAS) on pavement performance. A cost analysis was conducted to assess the lifecycle cost of asphalt pavements constructed with RAS. Then, a mix with 5% post-consumer waste shingles and recycled asphalt had the lowest cost over the pavement's service life. Nasim [36] investigated axial compression behavior of concrete columns concerning four types of aggregates (NA, RCA, RAP, and RCA-RAP). Reclaimed aggregate replacement ratios from 20% to 100% also were considered. Test results indicated that the ultimate failure load had a decreasing trend with the increase of the reclaimed aggregate substitute ratio, but the structure's safety was proved.

This study's main objective was to investigate the mechanical performance of RAP concrete, and several mechanical and physical laboratory tests were selected to evaluate its road performance in a cold region. Based on the various cement dosages, from 3.5% to 5.5% in the RAP concrete mix design, three recycled aggregate mixture contents (30%, 40%, and 50%) were selected to study the variation of mechanical behavior concerning different curing times, and the optimal RAP aggregate-substitute ratio was determined. A scanning electron microscope (SEM) was used to observe the inner-structure interface

between the asphalt binder and cement stone. A numerical model was created to simulate the RAP's compressive strength with respect to the effect of multiple parameters. The research results can provide a technical reference for RAP use in the reconstruction and expansion of low-grade highway projects.

#### **2. Experimental Programs**

#### *2.1. Reclaimed Asphalt Mixture Sieve Analysis and Mix Design*

Typically, asphalt pavement needs a high temperature for constant construction, and a thermal aging action occurs. In addition, the aging effect will be accelerated by light, weathering, snow/ice cover, rainwater penetration, and traffic load. In this paper, the reclaimed asphalt mixture came from the reconstruction and expansion of the National Dana Highway. The pavement structure was built in 2005, and consisted of 10 cm-thick asphalt concrete pavement. Due to the aging effect, the mechanical performance of the existing asphalt pavement cannot meet the serviceability. Therefore, the more extensive mixture was broken manually and then crushed with a small jaw crusher. Figure 1a,b show the RAP coarse and fine mixtures after being crushed by the jaw crusher, and some of them are clustered structures because of asphalt bonding. In addition, it can be observed that the asphalt and aggregate were separated due to sunlight weathering and traffic load. In order to reduce the difference of gradation composition, the RAP aggregate was sieved with an asphalt mixture centrifugal-extraction apparatus, as shown in Figure 1c,d.

(**a**) (**b**)

**Figure 1.** Reclaimed asphalt mixture: (**a**) coarse aggregate; (**b**) fine aggregate; (**c**) untreated. (**d**) Extracted RAP particle-size distribution (PSD).

Table 1 shows the physical properties of the RAP and NA aggregates. The RAP aggregate exerts a higher void content and lowers apparent density, and it can be attributed to uneven surfaces and microcracks generated during the RAP crushing process. In addition, the existing asphalt and cement binder lead to the enlarging of the inner air void, eventually causing a noticeable increase in water absorption for RAP.


**Table 1.** Physical properties of aggregates.

Considered with the various cement dosages from 3.5% to 5.5%, three recycled aggregate mixture contents (30%, 40%, and 50%) were selected in the RAP concrete mix design. The F30-S3.5 specimen represents 30% RAP and 70% NA, and the cement dosage is 3.5%. A total of nine combination specimen types were used, and their optimum water content and maximum dry densities are shown in Table 2.

**Table 2.** RAP concrete mix design.


#### *2.2. Axial Compressive-Strength Test*

According to the Highway Engineering Inorganic Binder Stability Materials Code (JTG E51-2009), a cylinder size with 150 mm diameter by 150 mm depth was selected. Its compression test setup is shown in Figure 2. Two variable parameters were considered in the experimental test: a reclaimed aggregate replacement ratio from 30% to 50%, and cement dosage from 3.5% to 5.5%. The specimens' compressive strengths were compared with respect to various curing times, from 7 to 90 days.

**Figure 2.** Axial compressive-strength test.

*2.3. Splitting Tensile Strength Test*

According to Highway Inorganic Bond Stabilization Materials Splitting Test Method (JTG E51-2009 T0806-1994), specimens with different cement dosages and reclaimed asphalt

mixtures were selected for 28- and 90-day curing times. Figure 3 shows the splitting-test setup and one typical test specimen.

(**a**) (**b**)

**Figure 3.** Splitting tensile-strength test: (**a**) test setup; (**b**) test specimen.

Similar to the splitting-strength test, specimens with two different curing times (28 and 90 days) were selected to test their compressive resilience moduli, and the relationship between multi-stage loading and their deformation was considered, as shown in Equation (1):

$$\mathbf{E\_c = \frac{\text{ph}}{l}}\tag{1}$$

where Ec is compressive resilience modulus (MPa); p is unit pressure (MPa); h is the specimen height (mm); and *l* is the resilient deformation (mm). Figure 4 shows the compressiveresilience modulus test setup and one typical test specimen.

**Figure 4.** Compressive-resilience modulus test: (**a**) test setup; (**b**) test specimen.

#### *2.4. Dry- and Temperature-Shrinkage Tests*

According to Highway Inorganic Bond Stabilization Materials Splitting Test Method (T0854-2009), water-loss rate and dry-shrinkage strain were selected to investigate the ability of deformation resistance under the presence of pore water. The water-loss rate was calculated with the mass variation when subjected to immersion and dry conditions, as shown in Figure 5a. The dry-shrinkage strain was measured with a micrometer gauge, and the monitor period was divided into two interval times, which was recorded daily during the first week, and extended from 2 to up to 27 days.

(**a**) (**b**)

**Figure 5.** Shrinkage test: (**a**) water immersion; (**b**) temperature-controlled cabinet.

Similarly, the temperature-shrinkage strain was measured with a micrometer gauge, and a lab temperature-controlled cabinet was required. Based on the on-site monitored data in Huma town (China), temperatures of 14.9 ◦C~18.7 ◦C in summer and −30.2~−20.1 ◦C in winter are observed. Thus, the temperature-shrinkage measuring zone was determined to be from 20 ◦C to −30 ◦C with an interval of 10 ◦C, as shown in Figure 5b.
