Fatigue and Healing Characteristics of RAP Binder Blends
Fatigue and Healing Characteristics of RAP Binder Blends
Fatigue and Healing Characteristics of RAP Binder Blends
Abstract: Reclaimed asphalt pavement (RAP) mixes generally have good moisture and rutting resistance. However, the fatigue performance
of the RAP mixes is a matter of concern because the RAP binder is generally brittle because of age hardening. Asphalt mixes can heal and
recover from the fatigue damage caused by repeated wheel load application. The healing characteristics of the mix depend to some extent on
the flow characteristics as well as the chemical makeup of the binders. Linear amplitude sweep (LAS) testing was performed for evaluating
the fatigue performance of RAP binders. The fatigue and healing characteristics of different virgin-RAP binder blends were measured in an
oscillation fatigue test conducted with varying rest periods using dynamic shear rheometer. The slopes of the fatigue life versus rest period
plots were taken as indicators of the effect of rest period on fatigue life or healing indices (HI). Different rheological and chemical parameters
measured in terms of surface free energy and the indices calculated using Fourier transform infrared (FTIR) spectroscopy correlated well with
the healing potential of RAP binders. DOI: 10.1061/(ASCE)MT.1943-5533.0003284. © 2020 American Society of Civil Engineers.
Author keywords: Reclaimed asphalt pavement (RAP) binder blends; Binder fatigue parameter; Healing; Linear amplitude sweep (LAS);
Fourier transform infrared (FTIR) spectroscopy; Surface free energy.
ence on the healing capability (Shan et al. 2010; Tan et al. 2012; indicators of the VG30 bitumen, RAP binder, and RAP-virgin
Ling et al. 2012; Shan et al. 2013). Most of the studies discussed binder blends prepared corresponding to different RAP propor-
here in the context of healing of binders and mixes were carried out tions used in the mix were evaluated in terms of penetration,
on conventional binders and mixes. softening point, viscosity, oscillation test, and linear amplitude
The dependence of the performance of bituminous mixtures on sweep (LAS) test parameters.
the chemical properties of the binders has been an area of research • The healing characteristics of the binders were evaluated by
focus for quite some time. Changes in the chemical composition conducting time sweep tests with intermittent loading on the
of the binder can influence the rheological properties of the binders binder samples.
which, in turn, affect the pavement performance. The Fourier trans- • Surface free energies of the binders were estimated using the
form infrared spectroscopy (FTIR) technique is commonly em- contact angles measured with the help of a goniometer.
ployed to estimate the presence of different functional groups in • The functional groups present in the binder blends were identi-
the binder. Similarly, surface free energy of asphalt binders, which fied using FTIR spectroscopy.
can be used to explain the work of cohesion and adhesion (Cheng
et al. 2002; Huang et al. 2005), can be related to the fatigue crack-
ing and healing characteristics of asphalt binders and mixtures Materials Used
(Little and Bhasin 2006). Different methods such as Wilhelmy
plate, pendant drop, maximum bubble pressure method, and sessile The VG30 unmodified binder was sourced from the Haldia refin-
drop method (Howson 2011) were used for measuring of the sur- ery, which is located in West Bengal, India. The RAP binder was
face free energy of bitumen. The sessile drop method is considered extracted from the samples of RAP material collected from the
to be the simplest among these techniques (Ahmad 2011; Kakar Kulpi plant of National Highway 117 using the solvent extraction
et al. 2016; Yi et al. 2018). method [ASTM D2172 (ASTM 2011)]. The binder was recovered
Considering that very limited literature is available on the healing from the solvent by rotary evaporator method [ASTM D5404
characteristics of RAP mixes and that there is even less literature on (ASTM 2012)]. The properties of the RAP binder are given in
the healing behavior of RAP binders, it is necessary to investigate Table 1.
the effect of RAP content on the fatigue and healing characteristics Different RAP-virgin binder blends were prepared using VG30
of binders. Because the chemical characteristics of binder can have as the virgin binder. The proportions of the virgin and RAP binders
significant influence on the performance of the binders, it is also used for preparing different blends were selected corresponding to
important to examine the relationship between the chemical proper- the proportions of RAP considered for the preparation of RAP
ties of RAP binders and their fatigue and healing behavior. mixes, i.e., 15%, 25%, 35%, and 45% (percentage of the weight
It is essential to be able to estimate the fatigue performance of as- of the RAP material by weight of the total mix). The RAP binder
phalt mixes, including RAP mixes, accurately. For this, it is necessary to the total binder (R=T) ratios obtained for 15%, 25%, 35%, and
to identify the factors contributing to the fatigue performance and to 45% RAP mixes are 0.126, 0.198, 0.266, and 0.324, respectively.
quantify the effect of different parameters including healing. Informa- These proportions were considered for the preparation of the blends
tion about the extent of the influence of rest periods on the fatigue of the RAP and virgin binders. For preparation of the virgin-RAP
lives of conventional as well as RAP mixes will be useful in explain- binder blends, the proportions of the RAP and virgin (VG30) bind-
ing the differences in the fatigue behavior of different pavements sub- ers were selected as per the corresponding R=T ratios. The blends
jected to different traffic loading conditions (rest periods). It is also were prepared by mixing the extracted RAP binder and the virgin
useful to be able to identify and correlate the specific chemical param- binder (VG30) in required proportions at the mixing temperature
eters of the binders that affect the fatigue and healing characteristics determined for virgin (VG30) binder. Blending was done for 2 min
of binders because this will help to identify suitable combinations of by hand mixing.
virgin binders and/or rejuvenators to be used in the RAP mixes. The VG30 virgin as well as the four virgin-RAP binder blends
The present study envisages examining these issues as a step were tested for different basic properties such as softening point
forward to enhance the understanding about the fatigue and healing and penetration and for different mechanical and chemical proper-
characteristics of RAP binder blends. The study also aims to pro- ties. The tests conducted on the binders include oscillation, visc-
vide insight into the dependence of the fatigue and healing behavior ometry, and LAS tests. These tests were conducted using a DSR.
of RAP binder blends on the rheological and chemical properties of
the binders.
Table 1. Properties of RAP binder
Average binder Penetration Softening Dynamic viscosity
Research Objective and Scope of the Study content (%) value, d (mm) point (°C) at 60°C (P)
The general objective of the present study is to evaluate the fatigue 3.96 (by weight 15 82 346,000
of mix)
and healing characteristics of RAP binder blends. The specific
The results of the preliminary tests conducted on the binders, re- light for the wavelengths of infrared light (400–4,000 cm−1 ). The
ported previously by Mullapudi et al. (2019) are given in Table 2. spectra were analyzed to obtain information on the changes in the
As expected, the binder became stiffer (reduction in the penetra- carbonyl groups occurring due to the addition of RAP binder.
tion and increase in the softening point and viscosity) with increases Reaction of the carbon atoms in the binder with oxygen produces
in the RAP binder proportion in the virgin-RAP binder blend. carbonyls (ketones) during the process of aging. ICO is the index
that indicates ageing due to the formation of ketones. The Aliphatic
index (ICH), which represents the proportion of aliphatics in the
Experimental Investigations binder, was found to reduce with increase in RAP content. This is
in agreement with the findings of Jung (2006), Araújo et al. (2011),
Yao et al. (2013), and Gong et al. (2017). ICO and ICH values were
Fourier Transform Infrared Spectroscopy
calculated using Eqs. (1) and (2) (Liu et al. 1998; Singh and Sawant
A FTIR NEXUS-870 spectrometer (Thermo Fisher Scientific, 2016). The ICO and ICH indices have been correlated with healing
Waltham, Massachusetts) was used in this study for the FTIR spec- index measured from linear amplitude sweep fatigue test conducted
troscopy. In this, the path of a light beam is interrupted by the sample. on the binders. Fig. 1 shows the FTIR spectrum obtained for the RAP-
The detector of the spectrometer measures the amount of absorbed virgin binder blend prepared corresponding to 45% RAP content.
The ICO and ICH values calculated from the FTIR spectra ob- The methylene and methyl hydrocarbon (MMHC) ratio, which
tained for different binder blends are presented in Table 3. is the ratio of hydrogen to carbon atoms in methyl and methylene
It is observed that ketones increased with increase in the pro- attached to aliphatic independent molecules or aliphatic chains
portion of RAP binder in the blends, indicating a stiffening effect associated with cycloalkanes or aromatic centers, represents
(Zofka et al. 2012). Aliphatics, which are the lowest molecular the branching of alkane material (Zeiada 2012). MMHC and
weight compounds (among the different fractions in the bitumen) CH2 =CH3 ratios were calculated using the absorbance values ob-
aromatize and thus are converted to polyaromatic carbon ring units tained from the infrared spectroscopy (Kim et al. 1990). The val-
(asphaltenes fraction). Hence, ICH decreased with increase in RAP ues of MMHC ratio for different RAP binder blends are presented
content. in Table 4.
Branching of the molecules is signified by MMHC ratio, and
chain length is indicated by CH2 =CH3. It is observed that MMHC
increased and CH2 =CH3 decreased with increase in RAP content.
Surface free
Binder energy (mJ=m2 )
Table 6. Results from LAS test
0% RAP 21.18
25% RAP 18.81 N f at Standard N f at Standard
45% RAP 18.18 Binder A B 1% strain deviation 2.5% strain deviation
Source: Data from Mullapudi and Sudhakar Reddy (2018). 0% RAP 402,112 4.66 402,112 172,416 5,403 1,502
15% RAP 77,995 3.93 77,995 48,615 2,133 1,358
25% RAP 18,774 4.23 18,774 5,204 382 26
35% RAP 5,198 6.35 5,198 781 17 3
Surface Free Energy of Binders
45% RAP 3,002 6.70 3,002 1,533 7 4
Surface free energy (SFE) is the work to be done in order to in-
crease the size of the surface of a phase (Kakar et al. 2019). The
sessile drop method was used in the present study for measuring the
SFE of the binders. For this, the contact angle between the binder and 13,200 kPa, respectively. The variation in the binder fatigue
surface and the probe liquid was measured using different probe parameter G sin δ with variation in the RAP binder proportion
liquids. The SFEs (polar and dispersive) were calculated using the (R=T ratio) is presented in Fig. 2.
measured contact angles (Mullapudi and Sudhakar Reddy 2018). It is observed from Fig. 2 that the G sin δ value increased with
The results are given in Table 5. increase in the RAP content in the binder blend. In addition, the
The results in Table 5 indicate that the SFE value reduced with log of G sin δ varied linearly with R=T ratio. The fact that the log-
increase in the RAP content. This trend is similar to the findings of arithm of G sin δ varies linearly with R=T ratio corroborates the
Cheng et al. (2002) and Wei and Zhang (2010), who reported that assumption widely made by different agencies including the Asphalt
surface energies reduced with aging. Institute (2014) in the binder blending exercise carried out for RAP
mix design.
Fatigue Characterization of RAP Binder Blends
Linear Amplitude Sweep Test
Oscillation and LAS tests were conducted on the binders to evalu-
ate their fatigue characteristics. A DSR was used for this purpose. LAS testing is conducted to determine the resistance of asphalt
The fatigue characteristics of binders are generally evaluated at binders to fatigue failure. In this test, linearly increasing strain am-
intermediate pavement temperatures on long-term laboratory-aged plitudes are applied in order to accelerate the fatigue damage. The
binders. The oscillation (time sweep) test was conducted at LAS test includes a frequency sweep test followed by a linear am-
10 rad=s angular velocity at a test temperature of 25°C. The LAS plitude strain sweep test, conducted to obtain the inputs required for
test was also conducted at 25°C temperature in accordance with the estimation of the fatigue damage using viscoelastic continuum
AASHTO TP101 (AASHTO 2014). damage (VECD) approach. The test was conducted using the DSR
The binder blends prepared as per the selected R=T ratios were equipment with 8-mm-diameter parallel plate geometry. PAV-aged
short-term aged at 163°C for 85 min using a rolling thin-film oven binder film of 2-mm thickness was tested. As a part of the estima-
according to ASTM D2872 (ASTM 2019) and were then long-term tion of the fatigue life of the asphalt binders, a frequency sweep test
aged in a pressure aging vessel (PAV) to produce accelerated long- was carried out at 25°C for the 12 frequencies specified in
term aging by means of pressurized air and elevated temperature. AASHTO TP101 (AASHTO 2014) at a constant strain level of
The binders were placed in the PAV for a duration of 20 h at a 0.1%. The amplitude sweep test was then carried out at the same
temperature of 100°C, after which the binders were degassed using test temperature in oscillatory shear mode with controlled strain and
a degassing setup according to ASTM D6521 (ASTM 2013). The at a frequency of 10 Hz. The amplitude of the strain was ramped
PAV-aged binders were used to conduct the time-sweep and linear linearly from 0% to 30% over 3,100 cycles (100 cycles per each
amplitude sweep tests. strain level). The slopes of the G versus frequency plots were de-
termined and used for the calculation of the damage parameter.
Binder Fatigue Parameter (G sin δ) Eq. (7) was used to estimate the fatigue life of the binders. The
constants A and B in Eq. (7) were calculated by following the steps
The Superpave binder fatigue parameter, G sin δ, was determined given for damage analysis in AASHTO TP101 (AASHTO 2014)
from the oscillation test conducted at 25°C on a 2-mm-thick binder
film using the 8-mm-diameter parallel plate arrangement of the dy- N f ¼ Aðγ max Þ−B ð7Þ
namic shear rheometer [ASTM D7175 (2015)]. The test was con-
ducted at a frequency of 10 rad=s in controlled strain mode with 1% where N f = binder fatigue life; A and B = fatigue model param-
strain. The G sin δ values of VG30, 15% RAP, 25% RAP, 35% eters; and γ max = maximum expected binder strain for a given pave-
RAP, and 45% RAP binder blends are 3,600, 5,800, 7,400, 9,900, ment structure, percent.
The constants A and B and the binder fatigue lives calculated Fig. 4. Variation of fatigue life from oscillation test with rest period.
from the data obtained from the LAS tests are presented in Table 6.
The LAS test showed that the number of cycles to fatigue failure
decreased with increase in RAP content.
Fig. 6. Relationship between LAS fatigue life and Superpave fatigue Fig. 10. Relationship between fatigue life from oscillation test and
parameter G sin δ. logðN f Þ from LAS test.
tigations in future with more R=T ratios and more sources of RAP 2018. “Self-healing properties of recycled asphalt mixtures containing
material. metal waste: An approach through microwave radiation heating.” J. En-
viron. Manage. 214 (May): 242–251. https://doi.org/10.1016/j.jenvman
.2018.03.001.
González, A., J. Valderrama, and J. Norambuena-Contreras. 2019. “Micro-
Data Availability Statement wave crack healing on conventional and modified asphalt mixtures with
different additives: An experimental approach.” Supplement, Road
All data, models, and code generated or used during the study ap- Mater. Pavement Des. 20 (S1): S149–S162. https://doi.org/10.1080
pear in the published article. /14680629.2019.1587493.
Guo, R. 2007. “Predicting in-service life of flexible pavements based on
accelerated pavement testing.” Ph.D. thesis, Dept. of Civil, Architec-
Acknowledgments tural, and Environmental Engineering, Univ. of Texas.
Howson, J. E. 2011. Vol. 73 of Relationship between surface free energy
The authors are grateful for the financial support by the Department and total work of fracture of asphalt binder and asphalt binder-
of Higher Education, Ministry of Human Resource Development aggregate interfaces. College Station, TX: Texas A & M Univ.
(MHRD), Government of India through research project “Future of Huang, S. C., R. E. Robertson, J. F. Branthaver, and J. Claine Petersen.
Cities” (Grant No. F. No. 4-22/2014-TS.I, dated January 23, 2014). 2005. “Impact of lime modification of asphalt and freeze–thaw cycling
on the asphalt–aggregate interaction and moisture resistance to moisture
damage.” J. Mater. Civ. Eng. 17 (6): 711–718. https://doi.org/10.1061
/(ASCE)0899-1561(2005)17:6(711).
References Jung, S. H. 2006. “The effects of asphalt binder oxidation on hot mix as-
phalt concrete mixture rheology and fatigue performance.” Ph.D. thesis,
AASHTO. 2014. Estimating damage tolerance of asphalt binders using the
Dept. of Chemical Engineering, Texas A&M Univ.
linear amplitude sweep. AASHTO TP101. Washington, DC: AASHTO.
Ahmad, N. 2011. “Asphalt mixture moisture sensitivity evaluation using Kakar, M. R., M. O. Hamzah, M. N. Akhtar, and J. M. Saleh. 2019. “Evalu-
surface energy parameters.” Ph.D. dissertation, Dept. of Civil Engineer- ating the surface free energy and moisture sensitivity of warm mix as-
ing, Univ. of Nottingham. phalt binders using dynamic contact angle.” Adv. Civ. Eng. 2019 (1): 1–
Al-Qadi, I. L., Q. Aurangzeb, S. H. Carpenter, W. J. Pine, and J. Trepanier. 15. https://doi.org/10.1155/2019/9153603.
2012. Impact of high RAP contents on structural and performance prop- Kakar, M. R., M. O. Hamzah, M. N. Akhtar, and D. Woodward. 2016. “Sur-
erties of asphalt mixtures. Urbana, IL: Illinois Center for Transportation. face free energy and moisture susceptibility evaluation of asphalt bind-
Araújo, M. F. A. S., V. F. C. Lins, and V. M. D. Pasa. 2011. “Effect of ageing ers modified with surfactant-based chemical additive.” J. Cleaner Prod.
on porosity of hot mix asphalt.” Braz. J. Pet. Gas 5 (1): 3275–3281. 112 (Jan): 2342–2353. https://doi.org/10.1016/j.jclepro.2015.10.101.
https://doi.org/10.5419/bjpg2011-0002. Kim, Y. R., D. N. Little, and F. C. Benson. 1990. “Chemical and mechani-
Asphalt Institute. 2014. Asphalt mix design methods. 7th ed. MS-2. cal evaluation on healing mechanism of asphalt concrete (with discus-
Lexington, KY: Asphalt Institute. sion).” J. Assoc. Asphalt Paving Technol. 59 (1): 240–275.
ASTM. 2011. Standard test methods for quantitative extraction of bitumen Kim, Y. R., D. N. Little, R. L. Lytton, J. D’Angelo, R. Davis, G. Rowe, P.
from bituminous paving mixtures. ASTM D2172. West Conshohocken, Ling, H. Jiang, S. Wu, and S. Wu. 2012. “Self-healing capacity of as-
PA: ASTM. phalt binders.” J. Wuhan Univ. Tech. Mater. Sci. Ed. 27 (4): 794–796.
ASTM. 2012. Standard practice for recovery of asphalt from solution Kim, Y. R., D. N. Little, R. L. Lytton, J. D’Angelo, R. Davis, G. Rowe, and
using the rotary evaporator. ASTM D5404. West Conshohocken, L. Tashman. 2002. “Use of dynamic mechanical analysis (DMA) to
PA: ASTM. evaluate the fatigue and healing potential of asphalt binders in sand as-
ASTM. 2013. Standard practice for accelerated aging of asphalt binder phalt mixtures.” In Vol. 71 of Proc., Technical Sessions-Asphalt Paving
using a pressurized aging vessel (PAV). ASTM D6521. West Technology: Association of Asphalt Paving Technologists, 176–206.
Conshohocken, PA: ASTM. Lino Lakes, MN: Association of Asphalt Paving Technologists.
ASTM. 2015. Standard test method for determining the rheological prop- Leegwater, G. A., A. Scarpas, and S. M. J. G. Erkens. 2018. “The influence of
erties of asphalt binder using a dynamic shear rheometer (DSR). boundary conditions on the healing of bitumen.” Road Mater. Pavement
ASTM D7175. West Conshohocken, PA: ASTM. Des. 19 (3): 571–580. https://doi.org/10.1080/14680629.2018.1418720.
ASTM. 2019. Standard test method for effect of heat and air on a moving Ling, P., H. Jiang, S. Wu, and S. Wu. 2012. “Self healing capacity of asphalt
film of asphalt (rolling thin film oven test). ASTM D2872. West binders.” J. Wuhan Univ. Tech. Mater. Sci. Ed. 27 (4): 794–796. https://
Conshohocken, PA: ASTM. doi.org/10.1007/s11595-012-0550-z.
Bahia, H. U., H. Zhai, K. Bonnetti, and S. Kose. 1999. “Non-linear visco- Little, D. N., and A. Bhasin. 2006. Using surface energy measurements
elastic and fatigue properties of asphalt binders.” J. Assoc. Asphalt to select materials for asphalt pavement. NCHRP Project No. 9-37.
Paving Technol. 68 (1): 1–34. Washington, DC: National Academic Press.
Bhasin, A., S. Palvadi, and D. N. Little. 2011. “Influence of aging and tem- Little, D. N., R. L. Lytton, and D. Williams. 1998. An analysis of the
perature on intrinsic healing of asphalt binders.” Transp. Res. Rec. mechanism of microdamage healing based on the application of micro-
2207 (1): 70–78. https://doi.org/10.3141/2207-10. mechanics first principles of fracture and healing. Lino Lakes, MN:
Bodin, D., H. Soenen, and C. de La Roche. 2004. “Temperature effects in Association of Asphalt Paving Technologists.
binder fatigue and healing tests.” In Proc., 3rd Eurasphalt and Little, D. N., R. L. Lytton, D. Williams, and C. W. Chen. 2001. “Microdam-
Eurobitume Congress. Vienna: Eurasphalt and Eurobitume Congress. age healing in asphalt and asphalt concrete.” In Vol. 1 of Microdamage