Effect of rejuvenators on the crack healing performance of recycled asphalt pavement by induction heating

Construction and Building Materials 164 (2018) 246–254

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Effect of rejuvenators on the crack healing performance of recycled asphalt pavement by induction heating Ba Huu Dinh, Dae-Wook Park ⇑, Tri Ho Minh Le Dept. of Civil Engineering, Kunsan National University, 558 Daehak ro, Kunsan, Jeonbuk 54150, Republic of Korea

h i g h l i g h t s
The optimum SWF content is determined by conductivity, and microstructure tests.
The induction heating method was processed on recycled asphalt mixes.
The conductivity increases with an increase in steel wool fiber content.
The healing performance is effective at a certain temperature.
The rejuvenators with low viscosity are applicable to RAP in the healing process.

a r t i c l e

i n f o

Article history: Received 19 August 2017 Received in revised form 6 December 2017 Accepted 27 December 2017

Keywords: Steel wool fibers CT-scan Induction heating Recycled asphalt pavement Rejuvenator Cooking oil waste

a b s t r a c t This paper evaluates the healing performance of recycled asphalt mixture modified with steel wool fibers (SWF) using induction heating method. The optimum SWF content was determined through conductivity and microstructure tests. The healing performance of SWF modified recycled asphalt mixture with different rejuvenators and cooking oil waste was evaluated. The testing results shows that the presence of RAP reduces the induction heating effectiveness due to the long-time oxidization and aging process. For the self-healing purpose, it was suggested that an addition of certain rejuvenator or cooking oil waste with low viscosity can enhances healing performance of the recycled asphalt mixture. Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction Asphalt mixture can repair its minor damages with self-healing ability [1]; however, the preparation process is very slow and incompletely under service condition, weather and repeated traffic loading [2]. Self-healing ability of bituminous mixtures showed to be a prevailing topic in recent years. Many methods have been investigated and developed to promote this potential phenomenon, such as nanoparticles, rejuvenator encapsulation, and induction healing. With the healing purpose, some researchers have recommended the employment of the steel wool fibers (SWF) into the asphalt mixture to enable the induction heating technology. These studies confirmed that the additive conductive plays an important role in healing performance of asphalt mixture. To promote the use of conductive asphalt mixtures for induction ⇑ Corresponding author. E-mail address: [email protected] (D.-W. Park). https://doi.org/10.1016/j.conbuildmat.2017.12.193 0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.

heating, many efforts were concentrated on adding fiber-type conductive particles (e.g. steel fibers, steel wool fibers, iron powder, carbon fibers). Vo [3] made an effort to improve the thermal properties of asphalt mixtures by using graphite and carbon fibers. The results presented that carbon fibers and graphite fundamentally enhance the thermal conductivity of asphalt mixtures. Garcia [4] evaluated the influence of conductive fibers and fillers on induction heating rate of asphalt mastic. The results showed that the fibers content and the ratio of sand–bitumen was directly related to the induction heating rate. Apostolidis [5] used steel fibers and iron powder to improve healing performance of asphalt mortar with induction heating. It was concluded that thermal, electrical conductivity and induction heating rate of asphalt mortar increases with increasing SWF content. They also found that the combination of iron powder and fibers exposed higher thermal conductivity than single additive. These findings indicated that the steel wool fibers with a small diameter and long length were a higher enhancement of electrical conductivity than the big diameter and

247

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254 Table 1 Particle size distribution of new aggregate. Sieve size (mm) Percent passing (%)

19 100

12.5 98

9.5 86

4.75 60

Table 2 Physical properties of asphalt binders. Properties

Standard

PG64-22

PG58-28

Flash Point, COC, °C Absolute Viscosity at 140°F (60 °C), Poises Penetration at 77 °F (25 °C), dmm Specific Gravity at 60 °F (15.6 °C)

T48 T202 T49 T228

265 2010 70 1.024

260 910 130 1.022

2.36 45

0.6 23

0.3 14

Evaluate the effect of rejuvenator on healing performance of recycle asphalt mixtures

Bitumen and steel wool fibers were heat at 170 oC in 2 h

RAP, bitumen and steel wool fibers were heat at 170 oC in 2 h

1 h before mixing, rejuvenator was added to asphalt and thoroughly stirred

Asphalt binder was mixed up with conductive additive in 60-120 secs to obtain well dispersion

The new binder was mixed with the aggregates in 90 secs at a constant temperature of 170 oC

After the mixing process, all materials were placed in the oven with a temperature of 140 oC in 2 h for aging

The mixtures were then compacted using a Superpave Gyratory compactor with the target air voids of 4 percent + 24 h curing at room condition

The compacted specimens were cut into 3 equal and separate portions.

0.075 3

short length fiber. Furthermore, they suggested that ten percent of the steel wool fibers by volume of asphalt binder is optimal content to acquire a high induction heating rate. Regards to the utilization of reclaimed asphalt pavement (RAP), this potential trend has become a huge attractive solution to many highway agencies around the world due to the conserve energy, environmental safety and sustainable development generated by this technology. Across some developed countries, RAP has been widely used as an alternated material into hot mixed asphalt (HMA) to shorten the original aggregate consumption and to

New aggregate was heat at 170 oC in 4 h before mixing

Determinate optimum steel wool fibers content

0.15 8

6 semi-circular samples were cut from the cylindrical specimens.

Fig. 1. Mixing process.

248

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254

mitigate the environmental effects of construction [6]. However, traffic service and weather have been caused hardened oxidation to the asphalt from RAP. When mixing with original asphalt binder, this deterioration will result in an asphalt binder that stiffer than the virgin one and thereby, reducing the HMA performance, fatigue life, and durability. Hence, many agencies still remain reluctant to use recycled asphalt pavement due to this serious challenge. Generally, rejuvenator is a type of asphalt additives containing a high amount of maltenes constituents which promote re-balance the composition of the aged binders that lost during traffic service. Many studies have highlighted the effectiveness of different rejuvenator types on the characteristic and the performance of the HMA. Tran [7] indicated that rejuvenators can soften the oxidized binder and enhance the cracking resistance of the mixtures. Zaumanis [8] evaluated the impact of various rejuvenators on the engineer characteristics of RAP binder and recycled asphalt mixtures. Nahar [9] determined the microstructure and rheological of HMA using rejuvenator. The author stated that original microstructure can be partially restored and new morphology can also be achieved by using rejuvenator. However, there is no report concerning the combined use of Rejuvenator and steel wool fibers in hot recycle asphalt mixtures and its healing effectiveness by induction heating method. This technology may be promising due to its cost-effective strategy and environmental safety. If this method is well generated, multi objectives will be obtained. It is expected that both Rejuvenator and RAP can be utilized in mixture with steel wool fibers since induction heating technology will help prevent the potential cracking issues.

Among some popular rejuvenators, researches on cooking oil waste (COW) are very promising. It is reported that the oil has been generated million tons per years. A huge amount of oil has caused serious environmental challenge because of its disposing into river and landfill. Meanwhile, only small fraction of this used oil is recycled in soaps, fuel diesel, or animal foods production. Recently, some researchers have proved the potential characteristics of cooking oil waste as rejuvenators in recycled asphalt pavement [10]. It is found that adding cooking oil waste into aged asphalt helped replenishes the volatiles and dispersing oils, restoring the aged binder rheological property. Hence, cooking oil waste will be employed as the main rejuvenator in this research. The objective of this study is to assess the influences of different rejuvenators on the self-healing performance of mixtures containing RAP and SWF in terms of several damage healing cycles. Another objective is to identify the healing performance between HMA using new aggregate and mixture containing high RAP content. For this purpose, asphalt mixture modified with different rejuvenator types, steel wool fibers, and RAP contents were developed. In this research, an optimum SWF dosage for induction heating was investigated. The thermal and electrical properties of asphalt mixture containing SWF were examined. CT-scan images were captured to observe the dispersion of SWF in the mixture. Based on MATLAB, an image analysis program was also developed as a new method to investigate the distribution of fibers within the asphalt mixture. Finally, for the effect of different rejuvenator types on healing performance of recycled asphalt mixture, induction heating and the three-point bending tests were conducted

Fig. 2. Thermal conductivity measurements: (a) Thermal Constants Analyzer, (b) Sensor set-up with specimens.

Fig. 3. Electrical resistivity measurements: (a) Resistance tester, (b) Digital multimeter.

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254

for eight healing cycles to evaluate the self-healing performance and stress-strain behavior.

2. Materials and methods 2.1. Materials The hot mix asphalt mixture was used in this research. The aggregates consisted coarse aggregate (grain size ranges from

249

4.75 to 12.5 mm), fine aggregate (grain size ranges from 0.075 to 4.75 mm), and filler (grain size was lower than 0.075 mm) with the gradation shown in Table 1. The bitumen used in this research was PG 64-22 and PG 58-28 for new and RAP asphalt mixtures respectively. The physical properties of asphalt binders were presented in Table 2. The additive for induction heating used in this study was steel wool fibers with an average diameter ranging from 70 to 130 lm, density of 7.18 g/cm3, length varied from 4 to 4.5 mm, and thermal conductivity of 80 W/m K. 2.2. Sample preparation The specimens were prepared according to the Superpave Mix Design Method [11]. The main mixing process is illustrated in Fig. 1. The mixtures were then compacted using a Superpave Gyratory compactor with the target air voids of 4 percent and a dimension of 64 mm in height and 100 mm in diameter. After 24 h curing at room condition, the compacted specimens were cut into 3 equal and separate portions for thermal conductivity and electrical resistivity measurement. 2.3. Thermal conductivity and electrical resistivity measurements The thermal properties of the asphalt mixtures with different SWF contents were measured to assess their thermal behaviors. The measurements on cylinder samples (20 mm height and 100 mm diameter) were conducted using a Thermal Constants Analyzer (model TPS 1500) (see Fig. 2). The measurement range of this tester was up to 5 MJ/m3 K of specific heat capacity and from 0.01 to 400 W/m K for thermal conductivity. The measurement process met ISO Standard 22,007-2 [12] and the accuracy of measurement was better than 5%. A sensor was placed between 2 specimens to measure their thermal properties. To ensure reliability for the experiment, there were total 18 measurements was done for 6 portions of each specimen. Electrical resistivity measurements were performed using a resistance tester (Fig. 3(a)) and digital multimeter (Fig. 3(b)) connected to two circular electrodes of aluminum with diameter 120 mm. The measurement ranges of resistance test were from 1 O to 20 MO for the digital multimeter and 0.01 MO to 10 GO for the resistance tester. The electrical resistivity q of asphalt mixture was calculated as Eq. (1)

Fig. 4. (a) Thermal conductivity and (b) electrical resistivity results.

q ¼ RS=L

Fig. 5. (a) CT scan original and (b) processed images.

ð1Þ

250

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254

Fiber area percentage (%)

0.9

SWF4

0.8

SWF6

Table 4 The chemical compounds of COW.

SWF8

0.7

Fatty acid

Content (%)

0.6

Lauric acid (C12:0) Myristic acid (C14:0) Palmitic acid (C16:0) Stearic acid (C18:0) Linoleic acid (C18:2n6c) Oleic acid (C18:2n9c) Linoleic acid (C18:3n3) Eicosenonic acid (C20:1) Behenic acid (C22:0)

0.48 1.25 12.68 4.58 52.74 21.22 6.42 0.37 0.26

Total

100

0.5 0.4 0.3 0.2 0.1 0.0

0

5

10

15

20

25

30

Height (mm) Fig. 6. The fibers area percentage of sample surface from bottom to top of specimen.

Table 5 Rejuvenator basic characteristics.

where R = the electrical resistance (O), L = 200 mm: the distance between two circular electrodes, and S = 7850 mm2: the surface area of the asphalt mixture specimen.

Type

Viscosity at 60 °C (cSt)

Specific gravity

R1 R2 R3 R4 COW

70–90 113 80 60 50–60

0.96 0.966 0.965 0.96 0.95

2.4. CT-scan To observe the dispersion of the fibers in mixtures, a thousand of cross-sectional images was obtained using a micro-CT scanner (SkyScan-1076). The scanner image can display the object with a small size of up to 0.35 mm, the visualization up to 4000  2300 pixels and three offset positions. The dimension of specimens was approximately 30  20  20 mm. Before testing, the specimens were heating with a temperature of 40 °C during 6 h to remove the moisture completely.

Table 6 Mixture combinations.

3. Determination of optimum SWF content 3.1. Thermal conductivity and electrical resistivity evaluation *

Fig. 4(a) shows the thermal conductivity values of the asphalt mixtures modified with SWF. The highest thermal conductivity values of asphalt mixture modified with 6% SWF was 20.47% higher than the control specimens. Regards to mixtures with SWF more than 6%, the fibers tend to tangle and form the bundles of fibers during the mixing process. This can cause uneven distribution of fibers in the mixture, thereby, thermal conductivity could be reduced. As the electrical conductivity is the reciprocal value of resistivity, r = 1/q, the resistivity is expected to be low for good conduction. The effect of SWF on the electrical resistivity of asphalt mixtures is shown in Fig. 4(b). In general, the electrical resistivity decreases as SWF content increases. When additive content reaches a certain level, approximately 6%, the resistivity reduced dramatically. A low content of SWF in asphalt stay as isolated clusters and are unable to form the continuous conductive path, hence the resistivity of the system has no significant change. The conductive network can be formed as a certain content level is reached. Beyond that level, the conductive network grows and spreads in all directions with the mixture. However, the further increase does not improve the conductivity significantly but probably affect the related properties of the asphalt mixture.

**

Mixture

Total asphalt (%*)

New asphalt (%*)

Rejuvenator (%**)

RAP (%*)

C R1-30 R1-40 R2-30 R3-30 R4-30 COW R30 R40

5.1 5.1 5.0 5.1 5.1 5.1 5.1 5.1 5.0

5.1 3.88 3.38 3.88 3.88 3.88 3.88 3.88 3.38

0 2.5 2.5 5.5 5.2 4.6 5 0 0

0 30 40 30 30 30 30 30 40

% by mass of total mixture. % by mass of new asphalt binder.

3.2. SWF dispersion analysis The CT-scan test was performed with the mixtures containing 4%, 6% and 8% of SWF to understand the distribution of fibers in the specimens. With each specimen, a thousand cross-sectional images were obtained from CT-scan results and arranged in order from bottom to the top of the sample. An image analysis program developed base on MATLAB (ver. 2013) was employed to calculate the fibers area percentage. Firstly, the original images (Fig. 5(a)) were converted to binary images (Fig. 5(b)) using a threshold method. The threshold method was interpreted as follows: Suppose the threshold value is T, if the pixel intensity value is greater than T, the pixel will be white; otherwise, it will be black. According to the observation, any non-black areas should be considered the fibers area; therefore, different global threshold levels were applied to images. In this case, the black and white pixels was discriminated with the threshold value of 0.55 and the program was identified the fibers areas as white pixels. The fiber area percentage was calculated by the ratio of white pixels to the total.

Table 3 Particle size distribution of RAP. Sieve size (mm) Percent Passing (%)

19 100

12.5 99

9.5 85

4.75 52

2.36 35

0.6 13

0.3 3

0.15 1

0.075 0

251

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254

In HMA production plant, the virgin RAP was crushed and screened to a nominal maximum aggregate size of 19 mm and this passing portion was used for this research. Based on AASHTO T308 with the ignition oven method, the measured binder content of the RAP was approximately 4.1%. Table 3 shows the RAP aggregate gradation that reached Superpave gradation requirements. Based on experiences, four different organic rejuvenators, numbered from 1 to 4, were chosen in this study by screening their popularity. The cooking oil waste (COW) collected from repeatedly fried oil from a local restaurant was additionally used as a softening additive for the RAP binder. The chemical compounds of used COW in this study were determined by the gas chromatographymass spectrometry (GC–MS) test provided by Kunsan National University Center for Research Facilities and the test results were shown in Table 4. The oil was filtered twice with tissue paper to remove the solid impurities which may cause bad diffusion into asphalt. The properties of the rejuvenators and oil provided by the manufacturer are briefly presented in Table 5.

Along the 30 mm height of the specimen, a total of 200 crosssectional images were selected for analysis and computation. The changes of fibers area percentage through the height of the specimens were shown in Fig. 6. Regards to mixtures with 4%, 6% and 8% SWF content, SWF4, SWF6, and SWF8, as the test specimens were scanned, the fibers area percentage showed a low range of variation, 0.17%–0.28%, 0.29%–0.44%, and 0.28%–0.82%, respectively. However, the range is relatively large in mixtures with 8% of SWF. For example, at a height of 6–9 mm, the average value of fiber area of SWF8 mixture was 0.45% while that value was 1.11% at a height of 15.3–18.3 mm. It indicates the fibers in this sample disperse unevenly. The higher fibers content was more easily entangled in the mixing process and therefore the bundles of fibers were formed. This result also demonstrates the decrease in thermal conductivity of mixture modified with 8% SWF shown in Fig. 5.

4. Healing performance of recycled asphalt mixtures 4.1. Materials’ characteristics

4.2. Mix design The characteristics of reclaimed asphalt pavement (RAP), rejuvenators and other modifications were described below. RAP was milled from pavements of various layers and locations in Korea.

Many research indicated that a hot recycling asphalt mixture with RAP content lower than 30% was confirmed to perform in a

100

C

R30

R40

R1-30

R1-40

Healing level (%)

80

60

40

20

0 1

2

3

4

5

6

7

8

No. of Cycles

(a) 100

C

R30

R40

R1-30

R1-40

Healing level (%)

80

60

40

20

0 1

2

3

4

5

6

7

No. of Cycles

(b) Fig. 7. Healing performance of recycled asphalt mixes at (a) 70 °C and (b) 90 °C.

8

252

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254

100

R1-30

R2-30

R3-30

R4-30

COW

Healing level (%)

80

60

40

20

0

1

2

3

4

5

6

7

8

No. of Cycles

(a) 100

R1-30

R2-30

R3-30

R4-30

COW

Healing level (%)

80

60

40

20

0 1

2

3

4

5

6

7

8

No. of Cycles

(b) Fig. 8. Comparing the effect of rejuvenators and cooking oil on healing performance at (a) 70 °C and (b) 90 °C.

manner equivalent to virgin hot mix asphalt [6,7]. Therefore, with the aim of determining the healing performance of blending RAP mixtures, two RAP contents, 30% (R30) and 40% (R40), were selected in this research. According to AASHTO M 323 [13], the asphalt binder used for RAP mixture was PG 58-28 with the physical properties presented in Table 2. The optimum asphalt binder used for mixture with 30% and 40% RAP was 5.1% and 5.0% respectively. The cooking oil waste and four different rejuvenator types were respectively mixed with 30% RAP, 6% steel wool fibers content. From preliminary experiments, the absolute viscosity value of the extracted asphalt binder from the RAP was approximately 30,000 Poise and this amount was used to design the following optimum Rejuvenator contents: 2.5, 5.5, 5.2, 4.6%, 5% by mass of new asphalt binder of mix R1-30, R2-30, R3-30, R4-30, COW respectively. The controlled specimens were also fabricated for later comparison. Table 6 describes the mix formulation, blending RAP ratios, optimum binder content and rejuvenator dosages. The RAP specimens were also fabricated in accordance with the Superpave Mix Design Method, followed as Fig. 1. The Superpave Gyratory compactor was used to compact specimens with the target air voids of 4 percent and a dimension of 115 mm in height and 150 mm in diameter since the designed nominal maximum size of

RAP were 19 mm. After 24 h curing at room temperature, the samples were cored to obtain the diameter of 100 mm based on the requirement of Three Point Bending (TPB) test. Then, 6 equal semi-circular samples were cut from the cylindrical specimens. Hence, the final size of the TPB test samples was 30 mm in height and 100 mm in diameter. Before conducting TPB test, a pre-existing crack was setup for each sample to control the initial position of the crack. Due to recommendations from preliminary papers and experiences, from the central axis of the loading and the mid-point of the semi-circular samples, a notch (pre-crack) was cut to obtain around 3 mm thickness and around 10 mm depth to prepared for the TPB test. Finally, these samples were cured at room temperature for about 48 h to completely dry out the moisture from the cutting process. 4.3. Healing processing The prepared samples were put in freezing chamber of the refrigerator (20 °C) for 2 h to acquire the brittleness condition for the initial Three Point Bending test. Semi-circular bending tests were carried out to determine the three-point bending strength of asphalt mixture samples. The TPB test setup contained a loading roller at the mid-point of the semicircular arch and space between

253

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254

Sh ¼

Fa F0

ð2Þ

where F0: the maximum force of the initially tested sample, and Fa: the maximum force of the sample after the healing process.

the improvements are relatively close; however, it can be clearly seen that the mixes with the rejuvenator types 1, 4, and the oil, R1-30, R4-30, and COW, are dominant in healing performance with a high healing level after the first cycle, especially mix COW. Whereas, mixes R2-30 and R3-30 have the healing level significantly decrease by about 40% at the second cycle. That might be due to the high viscosity of the rejuvenator (as detailed in Table 3) and therefore, causes slow and inefficient diffusion into the asphalt. Besides, after several healing cycles, mix COW still remains at high healing level, almost as much as that of the controlled mixture without RAP, or even better at a few late cycles as compared between Figs. 7(a) and 8(a). Fig. 9 shows the stress-strain behavior of mixes control, R30, COW at the initial state, first, fourth, eighth heating cycle. After reaching the peak stress, all the curves yield ductile behavior except at the last cycle as seen in Fig. 9(a). The mixes are brittle and no longer healed as the maximum stress is followed by an abrupt failure. In Fig. 9(b), with a replacement by 30% RAP, it is obvious that the induction heating still provides healing signal after the first cycle; however, compared to controlled mix, the

6000 5000

Stress (N)

two supporting rollers at the bottom edge was 8 cm. The universal testing machine used in this experiment has a capacity of 100 kN and the loading rate was 0.5 mm/min. The test was carried out at room temperature at around 23 °C. Each test sample was performed under TPB procedure until failure and a hairline crack was observed. After performing the TPB test, napkins were used to cover test samples and they were kept in ambient condition for 3 h to ensure that the moisture caused by the freezing process was totally dried out; then, the damaged sample was placed in induction heater and heated to reach desired healing temperature. The induction heating test was conducted by using an induction heating generator with a maximum frequency of 35 kHz and a controlled power of about 700 W. During the induction heating, the surface temperature of specimens was measured by an infrared camera with a period of 10 s until reaching the target temperature. After finished this step, the healed samples were let to rest for approximately 3 more hours at room temperature. This step will help the samples cool down and restore the stable condition. Then, the healed samples were preconditioned in freezing chamber one more time for 2 h before conducting the later TPB test. Finally, the healed samples were re-tested under TPB method and thereby finishing a healing cycle. In order to identify of the effectiveness of the healing process, eight damage-healing cycles were executed in the test samples. The healing level of asphalt mixture sample, Sh, was calculated by the Eq. (2) below:

4000 3000 Initial state

2000

First cycle Fourth cycle

1000

Eighth cycle

0

4.4. Healing performance

0.5

1

Strain

1.5

2

2.5

(a) 6000

Stress (N)

5000 4000 3000

Initial state

2000

First cycle Fourth cycle

1000

Eighth cycle

0

0

0.5

1

1.5

2

2.5

Strain

(b) 6000 5000 4000

Stress (N)

Fig. 7 shows the healing performance of recycled asphalt mixes, 30%, and 40% RAP, with the rejuvenators and an addition of 6% SWF. In general, the healing levels of all mixes decrease at every heating cycle. As compared between Fig. 7(a) and (b), the induction heating method is more effective as the heat generated to a higher temperature, 90 °C; this temperature is sufficient to heal the cracks without impacting on the mixture’s structure. The asphalt binder would be overflowed as if the heating temperature exceeds this threshold. The mixes with recycled asphalt have a low healing level and the healing level drops more than 50% at the first cycle compared to the controlled with only about 20% decrease after 6 cycles as seen in Fig. 7(b). The induction heating works well with the new asphalt but seems to be ineffective with RAP because its aging and oxidization lead to the high viscosity, which causes the low flowability to close the cracks. Therefore, the higher the RAP content in the mixture, the lower the healing level is reached. To address this problem, the rejuvenators were applied to the recycled mixes and they obviously enhance the healing performance. This helps soften the oxidized binder by retrieving the oils lost during age hardening and therefore, restores rheological properties of the RAP binder. In addition, the effect of the rejuvenator is more obvious at the generated temperature 90 °C compared to that at 70 °C as seen in Fig. 7(a). However, with the decrease by 10–30% at every cycle, the healing level of recycled mix could not reach to that of the controlled mix. The effect of rejuvenator types on the healing performance of recycled mix is shown in Fig. 8. All the rejuvenators have a certain improvement in the healing performance. However, as seen in Fig. 8(a), mixes R4-30 and COW show a good healing performance at the lower heating temperature. In Fig. 8(b), at the first cycle, all

0

3000 Initial state

2000

First cycle Fourth cycle

1000

Eighth cycle

0

0

0.5

1

Strain

1.5

2

(c) Fig. 9. Stress-strain curves of mixes (a) C, (b) R30, and (c) COW.

2.5

254

B.H. Dinh et al. / Construction and Building Materials 164 (2018) 246–254

healing performance is significantly low; mix R30 became brittle and seemed to be immune to the healing effect.

accompanied by a great improvement in healing performance together with a ductility behavior even at late cycles.

5. Conclusions

Acknowledgments

This study conducted several tests to promote the healing performance with induction heating method and apply to recycled asphalt pavements. The microstructure and the dispersion of fiber in the mixture were observed and analyzed. At two heating temperatures, eight healing cycles were conducted to recycled asphalt mixtures with different softening additives. The thermal and electrical conductivity of asphalt mixtures can be improved by adding SWF with a certain amount. In general, the conductivity increases as increasing SWF content. The entanglement of SWF causes the bad distribution of SWF in the mixture due to poor mixing and/or excessiveness. A large variation in the fibers area percentage through the height of specimen indicates the bad dispersion as the SWF content exceeds 6%. The surface temperature of asphalt mixture containing SWF generates higher with an increase in the fiber content and heating time. Based on the results, it is suggested that 6% SWF by volume of asphalt binder is reasonable for applying to recycled asphalt mixes using induction heating method. The 6% SWF in the asphalt mixture provided an optimum healing result as expected from the above suggestion. The healing performance of recycled asphalt mix significantly decreases at every heating cycle. The presence of RAP causes the ineffectiveness of induction heating since it is oxidized and aged after the longtime service. An addition of certain rejuvenators can enhance the healing performance of the recycled asphalt mix by reducing the RAP binder’s viscosity; however, the enhancement cannot provide the same healing level as the new asphalt. Additionally, the use of the softening additives is more effective at a sufficiently high heating temperature. The cooking oil waste and rejuvenators with low viscosity are suggested to apply to recycled asphalt pavement for self-healing purpose using induction heating method. Among four rejuvenators used in this study, mixture R4-30 has the healing performance stay at the highest level after several heating cycles. Especially, it is also revealed that the replacement of conventional rejuvenator by cooking oil waste in RAP mixture was apparently

This research was supported by a grant from Infrastructure and Transportation Technology Promotion Research Program funded by the Ministry of Land, Infrastructure and Transport of Korean Government (Code 17CTAP-C114670-02). References [1] A. Garcia, E. Schlangen, M. Ven, Q. Liu, Electrical conductivity of asphalt mortar containing conductive fibers and fillers, Constr. Build. Mater. 23 (2009) 3175– 3181. [2] A. Garcia, Self-healing of open cracks in asphalt mastic, Fuel 93 (2011) 264– 272. [3] H.V. Vo, D.W. Park, W.J. Seo, B.S. Yoo, Evaluation of asphalt mixture modified with graphite and carbon fibers for winter adaptation: thermal conductivity improvement, J. Mater. Civ. Eng. 29 (2017) 4016176–4016183. [4] A. García, E. Schlangen, M. Ven, Induction heating of mastic containing conductive fibers and fillers, Mater. Struct. 44 (2011) 499–508. [5] P. Apostolidis, X. Liu, T. Scarpas, G. van. Bochove, M.F.C. van de Ven, Advanced experimental evaluation of asphalt mortar for induction healing purposes, Constr. Build. Mater. 126 (2016) 9–25. [6] Recycling Hot Mix Asphalt Pavements. National Asphalt Pavement Association (NAPA), Information Series 123 Printed 4/96. [7] N.H. Tran, A. Taylor, Effect of Rejuvenator on Performance Properties of HMA Mixtures with High RAP and RAS Contents, National Center for Asphalt Technology, Alabama 12-05, 2012. [8] M. Zaumanis, R.B. Mallick, R. Frank, Evaluation of rejuvenator’s effectiveness with conventional mix testing for 100% reclaimed asphalt pavement mixtures, Proceedings of the XXVIII International Baltic Road Conference 2370, 2013. [9] S. Nahar, Turning back time: rheological and microstructural assessment of rejuvenated bitumen, Transp. Res. Rec. 2444 (2014) 52–62. [10] C. Meizhu, X. Feipeng, P. Bradley, L. Bingbing, W. Shaopeng, High-temperature properties of rejuvenating recovered binder with rejuvenator, waste cooking and cotton seed oils, Constr. Build. Mater. 59 (2014) 10–16. [11] Background of Superpave Asphalt Mixture Design and Analysis, Federal Highway Administration, Report Number FHWA-SA-95-003, 1995. [12] ISO Standard, Plastics – Determination of Thermal Conductivity and Thermal Diffusivity – Part 2: Transient Plane Heat Source Method, ISO Standard 220072, Switzerland, 2008. [13] American Association of State Highway and Transportation Officials, ‘‘AASHTO M 323: Standard Specification for Superpave Volumetric Mix Design”, Standard Specifications for Transportation Materials and Methods of Sampling and Testing, 30th Ed., AASHTO, Washington, DC, 2010.