Mechanical and Mesoscale Analyses of Cement Stabilized Macadam Prepared by Vibratory and Nonvibratory Mixing Techniques

-e objective of this study is to analyze the effects of mixing techniques on themechanical behavior andmeso-structure of cementtreated aggregate. Different specimens were prepared by vibratory and nonvibratory mixing techniques. X-ray CT scans were performed to illustrate the distribution of cement mortar on aggregate. -e strength, modulus, and fatigue tests under different stress states were tested to reveal the impacts of mixing techniques. -en, the relationships between strengths and loading rates and that between moduli and stress levels were established. Hereafter, the S-N fatigue equation that modified with stress ratio related to loading rates was used to describe the fatigue performance. -e results indicate that the cement mortar of specimens prepared by vibratory mixing was well-distributed on aggregates. -e strength, modulus, and fatigue life of the specimens prepared by vibratory mixing were higher under the test condition threshold. Moreover, the growth rate of strength and modulus with loading for specimens prepared by vibratory mixing was slightly larger than that for specimens prepared by nonvibratory mixing. Compared with the cement-treated aggregates specimens prepared by nonvibratory mixing, the fatigue life of cementtreated aggregates specimens prepared by vibratory mixing had more stable stress sensitivity.


Introduction
e semirigid asphalt pavement is one of the leading asphalt pavement types in China, taking advantage of strong bearing capacity, high integrity, and satisfactory capacity to distribute upper loads [1,2]. More than 90% of the bases of semirigid asphalt pavements in China use cement-treated aggregates [3,4]. Cement-treated aggregates are generally preferred materials for the semirigid base due to its high overall strength, high stiffness, good moisture susceptibility, desirable fatigue resistance performances, and readily available raw materials [5,6]. Cement-treated aggregates are usually prepared with aggregates of a suitable gradation and 3%-8% cement at the optimum water to cement ratio. e development of the strength of cement-treated aggregates is affected by the properties of constituent materials as well as the mixing technique [7][8][9]. e reflection cracks are the primary distress of semirigid base asphalt pavement [10,11]. In the literature, numerous studies were conducted to improve the pavement performance of cement-treated aggregates [12][13][14]. e influence of raw materials, cement content, and additives on the road performance of cement-treated aggregates was investigated. Microscopic analyses were also performed to optimize materials and design [15,16]. Ngoc Kien Bui et al. [17] proposed a new method that replaces natural aggregate (NA) with different percentages of recycled aggregate (RA). It was found that the combination of RA and NA could significantly improve the mechanical properties of recycled aggregate concrete (RAC) compared with the traditional method. Federico Autelitano et al. [18] replaced part of the aggregates in cement-treated aggregates with the electric arc furnace (EAF) slag. e authors stated that the compacted effect, durability, and mechanical properties of the modified mixture were significantly improved. . Eziefula et al. [19] developed a new green concrete, i.e., shell aggregate concrete, which could be used as nonstructural and low-strength ordinary concrete. All results of the above studies showed that the mechanical properties of cement-treated aggregates are closely related to the aggregate properties. William Fedrigo et al. [20] used indirect tensile tests, triaxial tests, and bending tests to evaluate the effects of cement content, curing time, and compressive strength on the strength and stiffness of cement-treated aggregates. e results showed that the cement content had the most significant influence on the strength and stiffness of cement-treated aggregates among the above factors. Xuan et al. [21] established a parameter prediction model to characterize the mechanical properties of cement-treated aggregates.
e authors reported that the unconfined compressive strength parameters could reflect the change of the mechanical properties of cement-treated aggregates. And, the unconfined compressive strength was mainly affected by the cement content, curing time, and degree of compaction. erefore, the mechanical properties of cement-treated aggregates are closely related to cement content, and the impacts of cement content, curing time, and compaction could not be ignored.
Besides, many scholars have shown that the type, structure, and parameters of mixers significantly affect the mixing uniformity, mixing efficiency, and hardening behavior of mixtures [22][23][24]. Vibratory mixing has become a popular technique for stirring concrete. Shen et al. [25] introduced different types of vibratory mixing techniques and gave the comparison among those mixing methods. Zhao et al. [26] used the vibratory mixing technique to form cement-treated aggregates specimens based on orthogonal test designs. It was found that the mixing time had the most noteworthy influence on the compressive strength of cement-treated aggregates, the vibration frequency and mixing speed was second, and the wet mixing time had the least impact [27]. Jiang et al. [28] compared the road performance of cement-treated aggregates prepared by the vibratory method and static method and showed that the vibratory method-prepared specimens were closer to engineering practice. Zhang et al. [29][30][31] conducted long-term experimental research on the vibratory mixing of cement-treated aggregates. ey stated that the vibratory mixing technique had a good wrapping effect on the coarse aggregate over traditional mixing technique so that cement-treated aggregates were more evenly mixed. And, they also found that the vibratory mixing technique could effectively improve the microscopic uniformity of cement-treated aggregates materials. us, this technique can promote a variety of dynamic effects of cement-treated aggregates materials, improve the structure formation process of concrete, and significantly improve the quality and efficiency of mixing. In the process of vibratory mixing, the mixing behavior realizes the macroscopic cyclic movement of the mixture, the purpose of which is to combine the materials; the vibratory behavior is used to accelerate the diffusion movement, the intention of which is to mix the materials evenly [25,32]. However, only a few researchers have revealed the influence mechanism of the mixing process on the road performance of cement-treated aggregates, and the research related to the influence of the mixing process on the mechanical properties of cement-treated aggregates materials is even insufficient. erefore, in this paper, the research has been conducted to verify the improving effects of vibratory mixing techniques from the perspective of mechanics and mesoscale structures.
e influence of mixing techniques on the pavement performance of materials was demonstrated. X-ray CT scans were carried to analyze the meso-structure of two different types of cement-treated aggregate specimens. e cement mortar and its wrapping effect under different mixing processes were also compared. en, the strength, modulus, and fatigue test of two different types of cementtreated aggregate specimens under different stress states were conducted under different loading rates. At the same time, the relationships between strength, modulus, and loading rates were established. e S-N fatigue equation associated with the loading rate was applied to characterize the fatigue properties of cement-treated aggregates.  Table 2. e gradation of aggregates is a crucial factor determining the mechanical properties of cement-treated aggregate. According to the Chinese Technical Guidelines for Construction of Highway Road bases (JTJ/T F20-2015) (2015), the gradation was selected based on the optimum service performance of the cement-treated aggregates; the gradation curve of limestone aggregates is shown in Figure 1.

Mixture Proportion Design.
e PSB32.5R cement with different contents (i.e., 3%, 3.5%, 4%, 4.5%, and 5%) and limestone aggregates with the target gradation ( Figure 1) was used to prepare cement-treated aggregates. Based on the requirements of Chinese Test Methods of Materials Stabilized with Inorganic Binders for Highway Engineering (JTG E51-2009) (2009), the vibration compaction test was applied to determine the maximum dry density and optimum moisture content of the mixture. e test results are presented in Table 3; the maximum dry density of cement-treated aggregates increased with the cement content, but when the cement content exceeded 4.5%, the increasing amplitude decreased obviously. us, the optimum cement content of 4.5% was calculated, and the corresponding optimum moisture content and maximum dry density were 4.8% and 2.429 g/cm 3 , respectively.

Specimen Preparation.
In order to analyze the influence of mixing techniques on the mechanical properties of cement-treated aggregates, the raw materials were mixed by the DETONG mixing equipment, which contained both vibratory and nonvibratory modes. And, the specimens were prepared using vibration compaction equipment.    Aggregates were sieved and then dried in an oven. After that, certain amounts of water, cement, and aggregate were mixed using the DETONG mixing equipment in the vibratory or nonvibratory mode for 40s. e vibratory frequency is determined based on the natural frequency of cementtreated aggregate for the purpose of resonance, and the amplitude is 0.5∼2 mm. And then, cement-treated aggregate specimens with two different mixing techniques were prepared in the mold under vibratory compaction.
According to the requirements of Chinese Test Methods of Materials Stabilized with Inorganic Binders for Highway Engineering (JTG E51-2009) (2009), beam specimens with sizes of 150 mm × 150 mm × 550 mm and cylindrical specimens with height of 150 mm and diameter of 150 mm were prepared. Among them, the beam specimens were prepared for four-point bending tests, and cylindrical specimens were prepared for unconfined compression tests and indirect tensile tests. After demolding, the specimens that met the requirements were stored in a standard curing room (the relative humidity of not less than 95% and the temperature of 20 ± 2°C) for 90 days, as shown in Figure 2.

Test Design.
In this study, X-ray CT was chosen to investigate the meso-structure of cement-treated aggregates prepared with different mixing techniques. e specimens were scanned with spacing of 0.9 mm at the voltage of 220 kV and the electric current of 10-40 mA. e strength, modulus, and fatigue tests were performed to examine the effect of mixing techniques on the mechanical performances of cement-treated aggregates. Indirect tensile tests were carried out using the Material Test System (MTS) equipment. Loading rates of 5 MPa/s, 10 MPa/s, 20 MPa/s, 30 MPa/s, 40 MPa/s, and 50 MPa/s were considered in the strength tests. Modulus tests were conducted using a staged loading method, and the stress levels were 0.25 MPa, 0.5 MPa, 1 MPa, and 1.5 MPa. e indirect tensile fatigue tests under stress-control mode and the halfsine load with a frequency of 10 Hz and the stress levels were 0.25 MPa, 0.5 MPa, 1 MPa, and 1.5 MPa. Unconfined compression tests and four-point bending tests were also carried out using the MTS equipment, and the test method was similar to that for the indirect tensile test. e loading rates of 5 MPa/s, 10 MPa/s, 20 MPa/s, 30 MPa/s, 40 MPa/s, and 50 MPa/s were also considered in unconfined compression tests and four-point bending tests. e stress levels used in four-point bending tests were consistent with those in indirect tensile tests. Since the loading rate for unconfined compression tests was too large, the stress levels were set at 3.5 MPa, 6 MPa, 8 MPa, and 10 MPa to ensure the safety and operability of the test. Five replicate samples were used for each test.

Effects of Mixing Techniques on the Strength.
Cement-treated aggregates specimens were prepared by vibratory and nonvibratory techniques. en, the specimens were tested by MTS under three different stress states to determine the strength. In this paper, the indirect tensile, unconfined compression tests, and four-point bending tests were chosen for the strength test.
In indirect tensile tests and unconfined compression tests, the loading rate was 1 mm/min. While the loading rate was 50 mm/min in four-point bending tests. e above choices were based on the Test Methods of Materials Stabilized with Inorganic Binders for Highway Engineering e test results are shown in Figure 3. It is noted from Figure 3 that the standard strength of the specimens prepared by vibratory mixing was higher than that of the specimens prepared by nonvibratory mixing. Under the standard test condition, the unconfined compressive strength (11.76 MPa) of the vibratory mixed specimen was about 18% larger than that (9.98 MPa) of nonvibratory mixed specimen; the indirect tensile strength (1.30 MPa) of the vibratory mixed specimen was about 15% higher than that (1.13 MPa) of nonvibratory mixed specimen; the bending tensile strength (1.31 MPa) of the vibratory mixed specimen was about 12% higher than that (1.17 MPa) of nonvibratory mixed specimen.
is indicates that the vibratory mixing has a more significant effect on the increase of unconfined compressive strengths of cement-treated aggregates.
In order to investigate the relationship between the strength and loading rate of specimens prepared by vibratory and nonvibratory mixing under three different stress states, different loading rates were applied in the tests. ree sets of parallel tests were carried out at the same loading rate, and the relationship curves between strength and loading rate of three different stress states are presented in Figures 4-6. Figures 4-6 show that the vibratory mixing had a more significant influence on the bending strength with the increased loading rate. As the loading rate continued to grow, both the unconfined compressive strength and the indirect tensile strength increase tended to be gentle.
Meanwhile, by comparing the parallel test strength results of the specimens prepared by vibratory and nonvibratory mixing, the coefficient of variation C v is shown in Table 4. It can be found that vibratory mixing can significantly reduce the coefficient of variation, which indicates that the cement-treated aggregates prepared by vibratory mixing had higher strength stability than the cement-treated aggregates prepared by nonvibratory mixing.

Effects of
where E c is the unconfined compressive modulus of the specimen (MPa); p is the unit pressure (MPa); h is the height of the specimen (mm); l is the rebound deformation of the specimen (mm): where E i is the indirect tensile modulus of the specimen (MPa); p is the load level (N); p 0 is the initial load (N); d is the diameter of the specimen (mm); and l x is the horizontal rebound deformation of the specimen (mm); μ is Poisson ratio: where E s is the four-point bending modulus of the specimen (MPa); p is the load level (N); p 0 is the minimum load (N); L is the span of the specimen (mm); l is the midspan rebound deformation (mm); bis the width of the midspan (mm); and h is the height of the midspan (mm). Figure 7 presents the standard modulus of vibratory and non-vibratory specimens under different stress states. It is noted that the standard modulus of the specimens prepared by vibratory mixing was higher than that of the specimens prepared by nonvibratory mixing. Under the standard test condition, the unconfined compressive modulus (16996 MPa) of the vibratory mixed specimen was about 22% larger than that (13913 MPa) of nonvibratory mixed specimen; the indirect tensile modulus (11507 MPa) of the vibratory mixed specimens was about 20% larger than that (9608 MPa) of nonvibratory mixed specimen; the bending tensile modulus (1976 MPa) of the vibratory mixed specimen was about 13% larger than that (1754 MPa) of nonvibratory mixed specimen. It was found that the vibratory mixing has a  Advances in Civil Engineering more significant influence on the increase of the standard modulus of cement-treated aggregates. In order to explore the relationship between the modulus and stress level of cement-treated aggregates specimens prepared by vibratory and nonvibratory mixing under three different stress states, different stress levels were applied in each stress state. And, the modulus tests were performed using a staged loading mode. Four sets of parallel specimens were tested at the same stress level. e relationships between the modulus and the stress level under three different stress states are illustrated in Figures 8-10. It can be noted that as the stress level increased under each stress state, the modulus decreased with a linear function. e linear function equation was as follows: where E is the modulus of the specimen (MPa); σ is the stress level (MPa); and k and m are the fitting parameters (see Table 5). Figures 8-10 show that, with the increase in stress level, the vibratory mixing had a more positive influence on the attenuation of the unconfined compressive modulus. As the stress level continued to grow, the mitigation of indirect tensile modulus and bending tensile modulus under vibratory conditions was also gradually accelerated, which was slightly better than the attenuation of the tensile modulus under nonvibratory conditions. e coefficient of variation C v of modulus results of specimens prepared by vibratory and nonvibratory mixing is shown in Table 6. It is noted that the coefficient of variation of the modulus of the specimen prepared by vibratory mixing was lower than that of the specimen prepared by nonvibratory mixing. In other words, the vibratory mixing           Figure 11.
It is observed that under standard conditions the fatigue life of the specimens prepared by vibratory mixing was higher than that of the specimens prepared by nonvibratory mixing (Figures 11 In order to investigate the variation of fatigue life of specimens at different stress levels, the unconfined compression tests, indirect tensile tests, and four-point bending fatigue tests were carried out regarding standard fatigue test procedures. e real stress ratio was obtained by the following equation: where t s is the stress ratio related to loading rates; σ is the stress level (MPa); and R c is the strength corresponding to the stress level (MPa). e fatigue life results and the real stress ratio were fitted by equation (6) in double logarithmic coordinates, as shown in Figures 12-14: where N f is the fatigue life of the specimen; t s is the real ratio; and a, k are the fitting parameters. Figures 12-14 show that the unconfined compressive, indirect tensile, and four-point bending fatigue life of the cement-treated aggregate varied significantly with the real stress ratio.
It is noted that the k value of the vibratory mixed specimen was significantly smaller than that of the nonvibratory mixed specimen.
is indicates that vibratory mixing was able to alleviate the fatigue failure of the cementtreated aggregates under any stress state. Advances in Civil Engineering e coefficients of variation C v of fatigue life results are shown in Table 6. It is found that vibratory mixing can significantly reduce the coefficient of variation, which indicates that the cement-treated aggregates of the vibratory mixed specimen had better stress sensitivity in terms of fatigue performance than that of the nonvibratory mixed specimen.

X-Ray CT Test Results and Analysis.
e cement-treated aggregate is a loose multiphase composite material, which has obvious nonuniform characteristics. Generally, it is easy to produce weak surfaces in the interior due to the uneven distribution of cementitious material and aggregate.
In order to study the dispersion of internal materials of cement-treated aggregates prepared by different mixing techniques, X-ray CT tests were carried out on the cylindrical specimens cured for 90 days. e X-ray CT test results of the circular cross section are shown in Figure 15.
It is observed that the specimens prepared by nonvibratory mixing had many voids between coarse aggregates. In other words, the voids were not filled by fine aggregates. By contrast, the specimens prepared by vibratory mixing had fewer voids and higher compactness. Fine aggregates in nonvibratory mixed specimens exhibited serious agglomeration. Because of the agglomeration of fine aggregates around coarse aggregates, coarse aggregates could be better wrap. Since the emergence of the phenomenon of agglomeration, the coarse aggregates cannot be fully wrapped.   Dense-skeleton structure Figure 15: e X-ray CT test results of the circular cross section of cement-treated aggregates prepared by different mixing techniques.

Conclusions
In this study, the strength, modulus, and fatigue life of cement-treated aggregates specimens prepared by different mixing techniques were tested. Meanwhile, the mesostructure of cement-treated aggregates was analyzed by X-ray CT. e following conclusions can be drawn from this study: (1) Compared with the nonvibratory mixing technique, the vibratory mixing technique enables cementtreated aggregates to have higher strength, modulus, and fatigue life. Moreover, the vibratory mixing techniques for improving unconfined compressive strength, modulus, and fatigue life were significantly higher than the other two stress state, which will facilitate its promotion and application. (2) Vibratory mixing technique significantly weakened the agglomeration of fine aggregates, so coarse aggregates could be sufficiently wrapped by fine aggregates. ereby the overall compactness of cement-treated aggregates was much improved by vibratory mixing.
(3) Further work could be focused on establishing the relationship between the microscopic characteristics and the mechanical properties of cement-treated aggregates.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest.