Effect of Nano Hydrotalcite on the Aging Resistance of a High Binder Content Stone Mastic Asphalt

Hydrotalcite, a type of layered double hydroxide (LDH), reveals an interesting potential for the modification of bitumen. The LDH can induce a barrier effect that prevents the loss of volatiles, retards oxidation, and protects against ultraviolet radiation. Such properties can enhance the aging resistance of the bitumen. However, there is a gap in knowledge regarding the effects of the modification with hydrotalcite in the properties of the asphalt mixture. To contribute to fill such a gap, the current study presents a characterization of the effects of the modification with nano hydrotalcite in the surface characteristics, mechanical performance, and aging resistance of an asphalt mixture. To better explore the effects of the modification, the selected asphalt mixture was a high binder content (7.5%) Stone Mastic Asphalt (SMA). The experimental study indicates that the binder-rich SMA presented adequate performance for application in surface courses. If compared to conventional mixtures, the binder-rich SMA presented better initial mechanical performance (unaged conditions). Furthermore, it presented smaller variation in the parameters between unaged and aged conditions, indicating enhanced aging resistance. The modification with nano hydrotalcite induced smaller evolution in the fatigue resistance parameters, indicating enhanced aging resistance; however, in the remaining tests, the trends were not clear.


Introduction
Asphalt pavements are a key element of the transport infrastructure, usually, being the dominant structure of the road network. During service life, pavements have to withstand a variety of traffic loads and environmental conditions (e.g., temperature variation, solar radiation, moisture, and oxidation) that will degrade their initial performance. Besides costly, the maintenance of these infrastructures is challenging due to the consequent disruptions in mobility and safety risks. Recent research in pavement engineering is focusing in improving the performance of the materials from economic, environmental and mechanical standpoints.
The surface courses are particularly important because of the direct contact with traffic. The asphalt mixtures applied in surface layer must perform well to provide adequate safety and comfort, coping with exigent requirements for skid resistance and permanent deformation, among others. Moreover, aging is particularly severe in pavement surface courses because they are directly exposed to the environmental conditions. Besides the conventional asphalt concrete (AC), over the years, other types of asphalt mixtures have been developed, for instance, the Stone Mastic Asphalt (SMA) and the Porous Asphalt (PA). In addition, to enhance the performance of these mixtures during the pavement service life, a variety of additives was studied.
In particular, SMA is a gap-graded asphalt mixture that exhibits good performance, especially for application in surface courses due to their high resistance to cracking and effect of the modification. An experimental program was developed to character asphalt mixtures, including testing for skid resistance, macrotexture, indirect strength, water sensitivity, stiffness, and resistance to fatigue. In addition, to asse sible long-term effects, such as aging resistance, the asphalt mixtures were condi for accelerated aging in laboratory, using the long-term oven aging (LTOA) metho tested afterwards. This paper presents and discusses the results of the experi program.

Materials
The asphalt mixture selected for the study was a stone mastic asphalt with mum size of aggregate 8 mm-SMA 8-designed according to the EN 13108-5 [3 ure 1 shows the particle size distribution of the aggregate mixture. Three aggrega tions, obtained from crushed gabbro rock, were used to achieve the final mixture gregates: 0/6, 2/6, and 6/12 (the numbers d/D represent the minimum and max nominal sizes of the aggregate fraction). The dosages by mass of aggregates we 19%, 54%, and 20% for the fractions of filler 0/6, 2/6, and 6/12, respectively. The asphalt mixture can be considered binder rich, as the selected binder c was 7.5% (by mass of asphalt mixture). The binder was a polymer-modified bi designated as PMB 45/80-65. In addition, 0.5% of natural cellulose fibers (by mass phalt mixture) were added to prevent the occurrence of binder drainage. The properties of the control asphalt mixture are presented in Table 1.
Two asphalt mixtures were produced: one designated as control (using 45/80-65 binder as supplied), and other designated as modified, where the PMB 45 binder was modified with 3% (by mass of final binder) of nano hydrotalcite. T drotalcite, also designated as magnesium aluminum hydroxycarbonate, has the f ing chemical formula: Mg6Al2CO3(OH)16·4H2O. The modification was done by heat original binder at 160 °C, adding the nano hydrotalcite powder and applying high stirring at 2000 rpm during 60 min. More details regarding the modification proce be found elsewhere [20]. Table 2 presents the physical properties of the contr modified binders. The asphalt mixture can be considered binder rich, as the selected binder content was 7.5% (by mass of asphalt mixture). The binder was a polymer-modified bitumen, designated as PMB 45/80-65. In addition, 0.5% of natural cellulose fibers (by mass of asphalt mixture) were added to prevent the occurrence of binder drainage. The main properties of the control asphalt mixture are presented in Table 1.
Two asphalt mixtures were produced: one designated as control (using PMB 45/80-65 binder as supplied), and other designated as modified, where the PMB 45/80-65 binder was modified with 3% (by mass of final binder) of nano hydrotalcite. The hydrotalcite, also designated as magnesium aluminum hydroxycarbonate, has the following chemical formula: Mg 6 Al 2 CO 3 (OH) 16 ·4H 2 O. The modification was done by heating the original binder at 160 • C, adding the nano hydrotalcite powder and applying high-speed stirring at 2000 rpm during 60 min. More details regarding the modification process can be found elsewhere [20]. Table 2 presents the physical properties of the control and modified binders.

Methods
The control and modified SMA 8 were produced and compacted in laboratory. Asphalt mixture production was conducted in a 30-L horizontal axis laboratory mixer, conforming to EN 12697-35 [40]. Cylindrical specimens (approximately with diameter of 101.6 mm and height of 63.5 mm) were compacted using the impact compactor, according to EN 12697-30 [41], applying 50 blows on each side. Prismatic specimens (approximately 400 mm × 300 mm × 50 mm) were compacted using the roller compactor, according to EN 12697-33 [42]. Prismatic specimens 400 mm × 50 mm × 50 mm were obtained by sawing the 400 mm × 300 mm × 50 mm specimens. Regarding test methods, Table 3 lists the parameters evaluated in the laboratory study. The surface characteristics of the control and modified asphalt mixtures were evaluated on prismatic specimens (300 mm × 400 mm × 50 mm) compacted using the roller compactor, as this compaction method is more similar to actual field compaction. The skid resistance was evaluated using the pendulum test [43] and the mean texture depth was evaluated using the volumetric patch technique [44].
Regarding mechanical performance, the control and modified SMA 8 were tested for indirect tensile strength, water sensitivity, stiffness and resistance to fatigue. To understand the effects of aging, for both asphalt mixtures-control and modified-a set of specimens was tested under unaged conditions and other set of specimens was tested after accelerated aging. Accelerated aging was conducted in laboratory according to the long-term oven aging (LTOA) method as described in AASHTO R30 [45]. In brief, the LTOA method consists in conditioning the compacted specimens inside a laboratory draft oven at 85 • C for 120 h. The aged asphalt mixtures were tested for indirect tensile strength, water sensitivity, stiffness, and resistance to fatigue.
The water sensitivity of the asphalt mixtures was determined through the indirect tensile strength ratio (ITSR), by testing the cylindrical specimens for indirect tensile strength (ITS). The ITS tests were conducted according to EN 12697-23 [46] and the water conditioning and ITSR determination according to EN 12697-12 [47]. The test temperature, 15 • C, was selected following the preconized by the Portuguese Infrastructure agency [48].
The stiffness evaluation was conducted according to EN 12697-26-Annex B [49], using the four-point bending (4PB) test. Under controlled temperature, the prismatic specimens (400 mm × 50 mm × 50 mm) were subjected to a sinusoidal controlled strain loading under the frequencies of 1, 3, 5, 10, 20, and 30 Hz. The selected strain level was 50 µm/m. To allow a broader understanding of the stiffness behavior, three test temperatures were selected: 10 • C, 20 • C, and 30 • C, respectively.
Resistance to fatigue was evaluated according to EN 12697-24-Annex D [50], using the 4PB test. The tests were performed at 20 • C, applying a 10 Hz sinusoidal loading under controlled strain. The adopted failure criterion was 50% reduction of the initial stiffness modulus.

Surface Characteristics
The surface characteristics of the control and modified asphalt mixtures were evaluated through the Pendulum Test Value (PTV) and the Mean Texture Depth (MTD). These characteristics are presented in Table 4. For each parameter, the average value and the respective standard deviation (in brackets) are presented. Both asphalt mixtures presented similar results, indicating that the modification with nano hydrotalcite had no particular influence in these properties. The Portuguese infrastructure agency [48] requires a PTV ≥ 60, however, indicating that the test can be conducted three months after the pavement is open to traffic. This recommendation aims to ensure that the binder film covering the aggregates was eliminated by traffic. In this study, the pendulum tests were conducted in fresh laboratory compacted specimens, with no binder film removal effect, thus the results obtained for the control and modified mixtures are considered adequate.
Regarding MTD, the Portuguese infrastructure agency [48] does not indicate limit values for SMA, but, for conventional asphalt concrete surface course (for instance, an AC The asphalt mixtures presented adequate performance in indirect tensile strength (ITS) and indirect tensile strength ratio (ITSR). The results of ITS and ITSR of the control and modified mixtures, under unaged and aged conditions, are presented in Figure 2. The error bars represent the respective standard deviation. ITS dry and ITS wet are the ITS of specimens under dry conditions (with no water sensitivity conditioning) and water sensitivity conditioned, respectively. Regarding MTD, the Portuguese infrastructure agency [48] does not indicate limit values for SMA, but, for conventional asphalt concrete surface course (for instance, an AC 10 surf) requires MTD ≥ 1.0 mm. Considering this criteria, both asphalt mixtures present adequate performance.

Indirect Tensile Strength and Water Sensitivity
The asphalt mixtures presented adequate performance in indirect tensile strength (ITS) and indirect tensile strength ratio (ITSR). The results of ITS and ITSR of the control and modified mixtures, under unaged and aged conditions, are presented in Figure 2. The error bars represent the respective standard deviation. ITSdry and ITSwet are the ITS of specimens under dry conditions (with no water sensitivity conditioning) and water sensitivity conditioned, respectively. However, in general, the differences are within standard deviation, for all aging and conditioning conditions the modified asphalt mixture presented higher ITS. Compared with the control material, the modified asphalt mixture presented 3% and 12% higher ITSdry and 9% and 11% higher ITSwet, for unaged and aged conditions, respectively.
As expected, the aging process caused an increase in ITS, that can be explained by the increase of the asphalt mixtures' brittleness. After aging, the control asphalt mixture presented 3% and 4% increase in ITSdry and ITSwet, respectively. In the case of the modified asphalt mixture, after aging the ITSdry and ITSwet increased by 12% and 6%, respectively. The effect of LTOA aging in the ITS of these SMA 8 seems less intensive if compared with conventional asphalt concrete (AC). For instance, in the study of Crucho et al. [51], that used the same test setup, the ITS of a conventional AC 14 mixture (with 4.5% of bitumen 35/50) evolved around 48%, from 1.29 MPa (unaged) to 1.91 MPa (LTOA aged). Other study [52], that tested (ITS tests at 20 °C) a Superpave mixture type SP-III with maximum aggregate size of 19 mm and 4.4% of bitumen PG 76-22, identified an increase of 28% in ITS after LTOA aging (unaged: 1.25 MPa; aged: 1.60 MPa). The high ITS values, as well as, the smaller effect of LTOA aging in these SMA 8 can be explained by the higher binder dosage and binder type. Regarding water sensitivity, the aging process had no clear effect in ITSR.

Stiffness
The results of the stiffness tests, namely, the stiffness modulus and phase angle, are presented in Figures 3 and 4, respectively. Under unaged conditions, in general, regardless of test temperature or loading frequency, the modified asphalt mixture presented lower stiffness modulus than the control asphalt mixture. Regarding phase angle, com- However, in general, the differences are within standard deviation, for all aging and conditioning conditions the modified asphalt mixture presented higher ITS. Compared with the control material, the modified asphalt mixture presented 3% and 12% higher ITS dry and 9% and 11% higher ITS wet , for unaged and aged conditions, respectively.
As expected, the aging process caused an increase in ITS, that can be explained by the increase of the asphalt mixtures' brittleness. After aging, the control asphalt mixture presented 3% and 4% increase in ITS dry and ITS wet , respectively. In the case of the modified asphalt mixture, after aging the ITS dry and ITS wet increased by 12% and 6%, respectively. The effect of LTOA aging in the ITS of these SMA 8 seems less intensive if compared with conventional asphalt concrete (AC). For instance, in the study of Crucho et al. [51], that used the same test setup, the ITS of a conventional AC 14 mixture (with 4.5% of bitumen 35/50) evolved around 48%, from 1.29 MPa (unaged) to 1.91 MPa (LTOA aged). Other study [52], that tested (ITS tests at 20 • C) a Superpave mixture type SP-III with maximum aggregate size of 19 mm and 4.4% of bitumen PG 76-22, identified an increase of 28% in ITS after LTOA aging (unaged: 1.25 MPa; aged: 1.60 MPa). The high ITS values, as well as, the smaller effect of LTOA aging in these SMA 8 can be explained by the higher binder dosage and binder type. Regarding water sensitivity, the aging process had no clear effect in ITSR.

Stiffness
The results of the stiffness tests, namely, the stiffness modulus and phase angle, are presented in Figures 3 and 4, respectively. Under unaged conditions, in general, regardless of test temperature or loading frequency, the modified asphalt mixture presented lower stiffness modulus than the control asphalt mixture. Regarding phase angle, compared with the control material, the modified asphalt mixture presented higher values for the test temperatures of 20 • C and 30 • C and lower values at 10 • C. Under aged conditions, the behavior was similar; however, the differences tended to be smaller. decrease of the phase angle. For instance, at 10 Hz and 20 °C, the LTOA aging caused a 16% and 24% increase in the stiffness modulus and 2% and 13% decrease in the phase angle, for the control and modified asphalt mixtures, respectively. In this regard, the binder-rich SMA was less sensitive to aging than a conventional AC (AC 14 with 4.5% of bitumen 35/50) tested using the same test setup and equipment [51]. The AC 14 presented 51% increase in stiffness and 31% reduction in phase angle (at 10 Hz and 20 °C) after LTOA aging.

Resistance to Fatigue
The fatigue lines, and respective coefficients of determination (R 2 ), of the control and modified asphalt mixtures under unaged and aged conditions are presented in Figure 5. Complementary results of the fatigue tests, namely, strain corresponding to the application of 100,000 loading cycles ( ) and the slope of the fatigue line (that is an indicator of the mixture's sensitivity to the applied loading strain level), are presented in Figure 6. After the aging process both asphalt mixtures, control and modified, presented worse resistance to fatigue. This behavior was expected, as the aging process causes hardening of the asphalt mixtures inducing a more brittle behavior. Nevertheless, the differences  decrease of the phase angle. For instance, at 10 Hz and 20 °C, the LTOA aging caused a 16% and 24% increase in the stiffness modulus and 2% and 13% decrease in the phase angle, for the control and modified asphalt mixtures, respectively. In this regard, the binder-rich SMA was less sensitive to aging than a conventional AC (AC 14 with 4.5% of bitumen 35/50) tested using the same test setup and equipment [51]. The AC 14 presented 51% increase in stiffness and 31% reduction in phase angle (at 10 Hz and 20 °C) after LTOA aging.

Resistance to Fatigue
The fatigue lines, and respective coefficients of determination (R 2 ), of the control and modified asphalt mixtures under unaged and aged conditions are presented in Figure 5. Complementary results of the fatigue tests, namely, strain corresponding to the application of 100,000 loading cycles ( ) and the slope of the fatigue line (that is an indicator of the mixture's sensitivity to the applied loading strain level), are presented in Figure 6. After the aging process both asphalt mixtures, control and modified, presented worse resistance to fatigue. This behavior was expected, as the aging process causes hardening of the asphalt mixtures inducing a more brittle behavior. Nevertheless, the differences  As expected, the aging process caused an increase of the stiffness modulus and a decrease of the phase angle. For instance, at 10 Hz and 20 • C, the LTOA aging caused a 16% and 24% increase in the stiffness modulus and 2% and 13% decrease in the phase angle, for the control and modified asphalt mixtures, respectively. In this regard, the binder-rich SMA was less sensitive to aging than a conventional AC (AC 14 with 4.5% of bitumen 35/50) tested using the same test setup and equipment [51]. The AC 14 presented 51% increase in stiffness and 31% reduction in phase angle (at 10 Hz and 20 • C) after LTOA aging.

Resistance to Fatigue
The fatigue lines, and respective coefficients of determination (R 2 ), of the control and modified asphalt mixtures under unaged and aged conditions are presented in Figure 5. Complementary results of the fatigue tests, namely, strain corresponding to the application of 100,000 loading cycles (ε 5 ) and the slope of the fatigue line (that is an indicator of the mixture's sensitivity to the applied loading strain level), are presented in Figure 6. After the aging process both asphalt mixtures, control and modified, presented worse resistance to fatigue. This behavior was expected, as the aging process causes hardening of the asphalt mixtures inducing a more brittle behavior. Nevertheless, the differences between unaged and aged asphalt mixtures were small, particularly if considering the dispersion of results commonly associated with the fatigue tests.
of −2% and −1%, in parameter and slope of the fatigue line, respectively. In the case of fatigue resistance, if compared with a conventional AC the difference between unaged and aged conditions was similar. Between unaged and LTOA aged conditions, an AC 14 with 4.5% of bitumen 35/50 tested using the same test setup and equipment [51], presented −4% of variation in the parameter ( of 244 µm/m and 235 µm/m, at unaged and aged conditions, respectively). However, if compared with the conventional AC 14, the fatigue resistance of the binder-rich SMA was superior.

Conclusions
This study is a contribution to fill a gap in knowledge regarding the effects of the modification with nano hydrotalcite in the properties of an asphalt mixture, particularly, surface characteristics, mechanical performance, and aging resistance. The asphalt mixture under study was a binder-rich Stone Mastic Asphalt with an aggregate upper sieve size of 8 mm. The used binder was a polymer-modified bitumen (PMB 45/80-65) and the binder dosage was 7.5% by mass. Two asphalt mixtures were studied, one using the supplied binder, and other using the supplied binder modified by adding 3% of nano hydrotalcite. To evaluate the performance of the asphalt mixtures a laboratory study was of −2% and −1%, in parameter and slope of the fatigue line, respectively. In the case of fatigue resistance, if compared with a conventional AC the difference between unaged and aged conditions was similar. Between unaged and LTOA aged conditions, an AC 14 with 4.5% of bitumen 35/50 tested using the same test setup and equipment [51], presented −4% of variation in the parameter ( of 244 µm/m and 235 µm/m, at unaged and aged conditions, respectively). However, if compared with the conventional AC 14, the fatigue resistance of the binder-rich SMA was superior.

Conclusions
This study is a contribution to fill a gap in knowledge regarding the effects of the modification with nano hydrotalcite in the properties of an asphalt mixture, particularly, surface characteristics, mechanical performance, and aging resistance. The asphalt mixture under study was a binder-rich Stone Mastic Asphalt with an aggregate upper sieve size of 8 mm. The used binder was a polymer-modified bitumen (PMB 45/80-65) and the binder dosage was 7.5% by mass. Two asphalt mixtures were studied, one using the supplied binder, and other using the supplied binder modified by adding 3% of nano hydrotalcite. To evaluate the performance of the asphalt mixtures a laboratory study was Under unaged conditions, compared with the control material, the modified asphalt mixture presented a 16% reduction in ε 5 and 13% increase in the slope of the fatigue line. However, the aging process caused higher variation in the parameters of the control asphalt mixture. Between unaged and aged conditions, the control asphalt mixture presented variations of −6% and 9% while the modified asphalt mixture presented variations of −2% and −1%, in parameter ε 5 and slope of the fatigue line, respectively.
In the case of fatigue resistance, if compared with a conventional AC the difference between unaged and aged conditions was similar. Between unaged and LTOA aged conditions, an AC 14 with 4.5% of bitumen 35/50 tested using the same test setup and equipment [51], presented −4% of variation in the parameter ε 5 (ε 5 of 244 µm/m and 235 µm/m, at unaged and aged conditions, respectively). However, if compared with the conventional AC 14, the fatigue resistance of the binder-rich SMA was superior.

Conclusions
This study is a contribution to fill a gap in knowledge regarding the effects of the modification with nano hydrotalcite in the properties of an asphalt mixture, particularly, surface characteristics, mechanical performance, and aging resistance. The asphalt mixture under study was a binder-rich Stone Mastic Asphalt with an aggregate upper sieve size of 8 mm. The used binder was a polymer-modified bitumen (PMB 45/80-65) and the binder dosage was 7.5% by mass. Two asphalt mixtures were studied, one using the supplied binder, and other using the supplied binder modified by adding 3% of nano hydrotalcite. To evaluate the performance of the asphalt mixtures a laboratory study was conducted, involving testing for skid resistance, mean texture depth, indirect tensile strength, water sensitivity, stiffness, and resistance to fatigue. To evaluate the effects of aging, both asphalt mixtures were aged under the LTOA method and subsequently tested for indirect tensile strength, water sensitivity, stiffness, and resistance to fatigue.
The obtained results indicated that the asphalt mixtures presented adequate surface characteristics and mechanical performance for application in surface courses. Regarding surface characteristics, the modified asphalt mixture presented approximately similar behavior to the control materials, having higher skid resistance and lower mean texture depth.
Regarding the initial mechanical performance (unaged conditions), the modified asphalt mixture presented a softer behavior, as indicated by a lower stiffness modulus. The most relevant enhancement of the modification was the increase in the indirect tensile strength and water sensitivity (with an ITSR of 100%).
The modification with nano hydrotalcite did not present clear contributions regarding a potential enhancement in aging resistance. To assess the effects of aging, the evolution of several mechanical performance parameters, between the unaged and aged conditions, was quantified. However the modified asphalt mixture presented smaller evolution in the fatigue resistance parameters, in the remaining tests the trends were not clear, or actually slightly worse than the control material.
The aged asphalt mixtures presented a more stiff and brittle behavior. As expected, after the aging process, the indirect tensile strength and stiffness modulus increased, and, the phase angle and resistance to fatigue decreased. If compared to conventional surface course asphalt mixtures, the binder-rich SMA tended to present better initial mechanical performance (under unaged conditions). Regarding aging, the binder-rich SMA presented smaller variation in the parameters between unaged and aged conditions, which indicates enhanced durability if compared to conventional asphalt mixtures.
Future studies will evaluate aging mechanisms under more field-like conditions, such as, the combined effects of temperature, solar radiation, and moisture damage.
The paper intended to give a contribution to the research and development of the bituminous mixtures nanomodification. It is a very promising technology regarding more sustainable and long life transportation infrastructures. The results presented in the paper have validated the techniques implemented in laboratory concerning the production and testing of this type of modified bituminous mixtures. However, some important challenges must be faced in near future to promote the generalized development and implementation of this technology in engineering practice. It is the case of supply form and cost of nanomaterials.
Author Contributions: Conceptualization, J.C. and J.N.; methodology, J.C. and J.N.; investigation, J.C. and J.N.; writing-original draft preparation, J.N.; writing-review and editing, J.C. and J.N.; supervision, J.C. and J.N. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.