Effect of Aging on Foamed Warm Mix Asphalt Produced

by Water Injection
Ala R. Abbas 1; Munir Nazzal, M.ASCE 2; Savas Kaya 3; Sunday Akinbowale 4; Bijay Subedi 5;
Mir Shahnewaz Arefin 6; and Lana Abu Qtaish 7
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.

Abstract: This study evaluated the aging characteristics of foamed warm mix asphalt (WMA) produced by water injection in comparison to
traditional hot mix asphalt (HMA). Two asphalt binders (PG 70-22 and PG 64-22) and one aggregate (12.5-mm NMAS limestone aggregate)
were used in this study. The short-term and long-term aging of the two asphalt binders were simulated using the rolling thin film oven (RTFO)
and pressure aging vessel (PAV), respectively, while AASHTO’s standard practice for mixture conditioning for hot mix asphalt was used to
simulate the short-term and long-term aging of the laboratory-prepared asphalt mixtures. The dynamic shear rheometer (DSR) was used to
characterize the viscoelastic behavior of the original (unaged, RTFO-aged, and PAV-aged) and recovered asphalt binders; Fourier-transform
infrared (FTIR) spectroscopy was used to identify and quantify the amount of functional groups present in the asphalt binders; and gel
permeation chromatography (GPC) was used to determine the molecular size distribution within the asphalt binders. In addition, this study
investigated the effect of the extraction and recovery procedures on the recovered binder properties. It was observed that the extraction and
recovery procedures had little effect on the rheological properties of PG 70-22, but a significant effect on PG 64-22. In addition, G
= sin δ and
G
sin δ values were obtained for the asphalt binders recovered from short-term and long-term oven-aged HMA mixtures that were
comparable to or slightly higher than those recovered from foamed WMA mixtures. This indicates that foamed WMA mixtures undergo
comparable or lower levels of aging than traditional HMA mixtures. The DSR test results were confirmed by the FTIR and GPC test results.
DOI: 10.1061/(ASCE)MT.1943-5533.0001617. © 2016 American Society of Civil Engineers.
Author keywords: Warm mix asphalt; Hot mix asphalt; Aging; Extraction and recovery; Dynamic shear rheometer; Fourier-transform
infrared spectroscopy; Gel permeation chromatography.

Background in 2008 to nearly 60% of the total amount of asphalt mixtures

produced in the state.
Warm mix asphalt (WMA) has become more widely adopted in the To date, satisfactory performance has been obtained for pave-
United States because of its environmental benefits, energy savings, ments constructed using foamed WMA, with minimal issues aris-
enhanced compaction, and increased haul distances. Over the last ing from the reduction of the production temperature. However, one
decade, different types of WMA technologies have been marketed subject that has not been thoroughly studied that might affect the
and used across the country. However, foamed WMA produced by performance and durability of foamed WMA is binder aging. Since
water injection has gained popularity among asphalt mix producers
lower temperatures are used during the production of foamed
in Ohio as it allows for the production of WMA with a standard
WMA, it is generally expected that the asphalt binders in these mix-
grade asphalt binder through a one-time mechanical plant modifi-
tures will undergo less aging, leading to lower resistance to perma-
cation, eliminating the need for costly additives associated with
nent deformation but better resistance to thermal and fatigue
other WMA technologies. In recent years, the amount of foamed
WMA used in Ohio has increased from approximately 10,000 tons cracking than traditional hot mix asphalt (HMA). However, the dif-
ference in aging between foamed WMA and HMA may also be
1
Associate Professor, Dept. of Civil Engineering, Univ. of Akron, affected by other factors such as the mix constituents. Therefore,
Akron, OH 44304 (corresponding author). E-mail: abbas@uakron.edu there is a need to investigate the aging characteristics of foamed
2
Associate Professor, Dept. of Civil Engineering, Ohio Univ., Athens, WMA mixtures to better understand their influence on pavement
OH 45701. performance.
3
Professor, Dept. of Electrical Engineering and Computer Science, Ohio
Univ., Athens, OH 45701.
4
Graduate Assistant, Dept. of Civil Engineering, Univ. of Akron,
Akron, OH 44304. Literature Review
5
Graduate Assistant, Dept. of Civil Engineering, Univ. of Akron,
Akron, OH 44304. Aging is a major factor affecting the life span of an asphalt
6
Graduate Assistant, Dept. of Civil Engineering, Univ. of Akron, pavement. Asphalt aging takes place during the production and
Akron, OH 44304. construction phase (short-term aging) as well as during the pave-
7
Graduate Assistant, Dept. of Civil Engineering, Ohio Univ., Athens, ment service life (long-term aging). Upon aging, the physical and
OH 45701. chemical properties of the asphalt binder change, causing it to
Note. This manuscript was submitted on July 25, 2015; approved on
become harder and more prone to cracking. Any cracks on the
February 12, 2016; published online on June 3, 2016. Discussion period
open until November 3, 2016; separate discussions must be submitted pavement surface may accelerate the aging process because of
for individual papers. This paper is part of the Journal of Materials in Civil the increased exposure to air and result in further pavement deterio-
Engineering, © ASCE, ISSN 0899-1561. ration, leading to premature pavement failure.

© ASCE 04016128-1 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128

 In recent years, several research studies have been conducted to
evaluate the aging characteristics of WMA binders and mixtures.
Gandhi and Amirkhanian (2008) studied binder aging in asphalt
mixtures prepared using two asphalt binders and two WMA addi-
tives (Asphamin and Sasobit). Asphalt binders extracted from
freshly prepared asphalt mixtures were considered to be short-term
aged, and asphalt binders extracted from oven-aged asphalt mix-
tures were considered to be long-term aged. The experimental test
results for viscosity, high- and low-temperature properties, and gel
permeation chromatography showed that asphalt binders extracted
from WMA had significantly lower aging as compared to binders
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.
extracted from the control HMA. It was also reported that the
WMA additives did not have any significant effect on the fatigue
cracking parameter (G
sin δ) or the creep stiffness of the asphalt
binders. However, Asphamin significantly increased the m-value
of the binders.
Arega et al. (2011) investigated the influence of warm-mix ad-
ditives and reduced aging on the rheology of asphalt binders with
different natural wax contents. Four asphalt binders (PG 76-22 and
PG 76-28 asphalt binders with high natural wax content and two
PG 64-22 binders with low natural wax content) and five WMA
additives (Evotherm DAT, Evotherm 3G, Sasobit, Rediset
WMX, and Cecabase RT 945) were included in the study. It
was reported that the cumulative effect of short-term aging fol-
lowed by PAV aging on asphalt binder stiffness depended on the
type of the binder and the WMA additive. It was also found that
certain WMA additives may reduce the viscosity of short-term aged
binders, especially those containing higher natural wax content.
This difference was more significant when asphalt binders were
subjected to a longer period of short-term aging.
Trujillo (2011) examined the rheological properties of a PG
64-22 asphalt binder blended with Cecabase RT, Rediset, Evo-
therm, and Sasobit WMA additives as a function of laboratory
aging. Control asphalt binder samples and binders modified with
WMA additives were aged in a rolling thin film oven (RTFO)
at 163 and 143°C, respectively. All samples were long-term aged
in an environmental chamber maintained at 60°C and subsequently
tested over a period of six months. Dynamic shear rheometer (DSR)
testing was performed on the aged samples at 45, 60, and 76°C
using a testing frequency of 0.1–25 Hz. Bending beam rheometer
(BBR) and Fourier-transform infrared (FTIR) spectroscopy were
also conducted on the aged samples. The FTIR results were re-
ported to show higher oxidation levels for the control samples than
the WMA samples, while the DSR and BBR test results showed
similar stiffness for the control and Sasobit samples and lower stiff-
ness for the Cecabase, Evotherm, and Rediset samples.
Ahmed et al. (2012) evaluated the effect of 15 warm mix
additives and dispersants on the rheological, aging, and failure
properties of four asphalt binders. The DSR test was conducted
on the unaged binder and RTFO residue to determine the high-
temperature grade. The PAV residue was tested in the DSR to
determine the intermediate grade and the BBR to determine the
low-temperature grade. In addition, a modified BBR test and the
double-edge-notched tension test were used to evaluate the low-
strain rheological and high-strain failure characteristics of the
asphalt binders. Significant changes in rheological and failure prop-
erties as well as asphalt binder grade span were reported as a result
of the addition of the warm mix additives and dispersants. The ad-
dition of the additives and dispersants was also reported to affect
the tendency to undergo chemical and physical hardening.
Banerjee et al. (2012) evaluated the effect of four warm mix
asphalt additives (Sasobit, Rediset, Cecabase, and Evotherm) on
the long-term aging characteristics on a PG 64-22 asphalt binder.
Shear testing of the control and WMA binders was conducted in the
© ASCE 04016128-2
J. Mater. Civ. Eng., 2016, 28(11): 04016128
laboratory at various levels of aging using the DSR test, and
statistical analysis was utilized to model the effect of temperature,
loading rate (or frequency), and aging on the asphalt binder
dynamic shear modulus. All WMA additives were observed to
reduce the dynamic shear modulus of the control asphalt binder.
It was reported that the Rediset WMA binder had the lowest shear
modulus, followed by the Evotherm, Cecabase, and Sasobit WMA
binders. It was also observed that the control PG 64-22 had the
highest rate of aging, while the Rediset WMA binder had the lowest
rate of aging, followed by the Evotherm, Cecabase, and Sasobit
WMA binders.
Punith et al. (2012) investigated the influence of long-term
aging on moisture susceptibility of foamed WMA mixtures con-
taining moist aggregate. Weight loss, indirect tensile strength
(ITS) of dry and conditioned specimens, and deformation (flow)
were measured for all mixtures. The experimental design included
two aggregate moisture contents (0 and ∼0.5% by weight of the dry
mass of the aggregate); two lime contents (1 and 2% lime by weight
of dry aggregate) and one liquid antistripping agent; one foaming
WMA additive (Asphamin) and two foaming water contents (2 and
3%); and two aggregate sources (granite and schist). It was reported
that the long-term aging improved the moisture resistance of the
WMA mixtures regardless of the antistripping agent and moisture
conditioning, with the aged WMA mixtures generally having a
greater wet ITS value than the aged control mixtures.
Xiao et al. (2012) conducted a study to examine the influence
of short-term aging on the rheological properties of nonfoaming
WMA binders. The experimental plan included four asphalt binders
and four nonfoaming WMA additives. Viscosity testing, perfor-
mance grading, creep and creep recovery, amplitude sweep, and
frequency sweep were performed to determine the influence of
the nonfoaming WMA additives on the asphalt binders. As ex-
pected, it was observed that the nonfoaming WMA additives
can reduce the viscosity of the asphalt binder and thus decrease
the mixing and compaction temperatures for the asphalt mixture.
A slight increase in failure temperatures was also reported for
the unaged binders and RTFO residues containing nonfoam
WMA additives as compared to the virgin asphalt binder. The
experimental test results also showed a slightly higher complex
modulus for the unaged binders and RTFO residues containing
Sasobit but lower creep compliance and phase angle than binders
containing other WMA additives.
Hasan et al. (2013) compared the performance of traditional
HMA mixtures to foamed WMA mixtures produced by water
injection and foamed WMA mixtures prepared using a foaming
additive (Advera). Four tests were included in the laboratory testing
program: asphalt pavement analyzer, flow number, indirect tensile
strength ratio, and four-point bending beam fatigue. It was reported
that the fatigue life of most foamed WMA samples prepared using
Advera was higher than that of the control HMA samples. How-
ever, foamed WMA hada significantly lower aging factor in com-
parison to HMA, which resulted in a higher rutting potential.
Kim et al. (2013) evaluated the short-term aging characteristics
of polymer-modified asphalt mixtures that incorporated two WMA
additives (Asphamin and Sasobit) using gel permeation chromatog-
raphy (GPC). The polymer-modified asphalt binders containing the
WMA additives were aged in the RTFO at 135 and 163°C for
85 min to simulate the short-term aging of the asphalt binder that
takes place during production, transportation, and construction.
Short-term asphalt mixture aging was simulated in the lab by
placing the loose asphalt mixture in a forced-draft oven for 2
and 4 h at 135°C and for 2 and 4 h at 154°C. The experimental
test results showed a higher level of aging for the asphalt mixture
short-term aging procedures than for the RTFO method, which can
J. Mater. Civ. Eng.
 be attributed to the thinner asphalt binder film thickness on the
aggregates than in the RTFO test. It was also reported that
the use of WMA additives resulted in lower binder aging for the
polymer-modified asphalt mixtures.
Hossain and Zaman (2013) evaluated the viscoelastic properties
of an asphalt binder containing different percentages of a
wax-based WMA additive, and utilized these properties to estimate
the dynamic modulus (E
) of the resulting WMA mixtures using
the Witczak and Hirsch models. The RTFO test was used to sim-
ulate the short-term aging of the asphalt binder. The Witczak
model, which is based on the DSR test results, was found to sig-
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.
nificantly underestimate the dynamic modulus, while the Hirsch
model was found to provide better approximations of the E
values.
In a research project funded by the NCHRP, West et al. (2014)
compared the material properties and field performance of pave-
ment sections constructed between 2006 and 2010 using WMA
and traditional HMA at 14 different locations across the United
States. Each of the 14 projects included at least one pavement
section constructed using WMA as well as a control section con-
structed using HMA. Several tests including the dynamic modulus,
flow number, AASHTO T 283, Hamburg wheel tracking test
(HWTT), uniaxial fatigue, and indirect tension compliance and
strength were performed in the lab on field-produced asphalt
mixtures. It was found that the WMA pavement sections had compa-
rable performance to those constructed using HMA in terms of
rutting, cracking, and moisture-induced damage, although a signifi-
cant difference in material properties was observed between the two
asphalt mixtures. In addition, it was reported that binders recovered
from WMA mixtures during construction had slightly lower stiffness
than those recovered from HMA mixtures in the DSR test. However,
binders recovered from WMA and HMA mixtures 1 to 2 years after
construction showed comparable results in the DSR test.
In another NCHRP project, Martin et al. (2014) evaluated the
performance of several pavement sections constructed at four sites.
The first site included three test sections constructed using tradi-
tional HMA, Evotherm 3G, and Sasobit. The second site included
four test sections constructed using traditional HMA, Evotherm
3G, Sasobit, and the Madsen Eco-Foam II foaming process. The
third site included three test sections constructed using HMA, Evo-
therm DAT, and foamed WMA. The fourth site included three test
sections constructed using HMA, Evotherm 3G, and foamed
WMA. Recycled asphalt pavement (RAP) was used in all test sec-
tions at the first and fourth sites, but not in the test sections at the
second or third sites. Several tests were used to characterize the
moisture susceptibility and long-term performance of the HMA
and WMA mixtures, including AASHTO T 283, HWTT, and
the resilient modulus test on unconditioned and conditioned spec-
imens. Several short-term oven aging procedures were used in the
preparation of the WMA specimens, which involved heating the
loose asphalt mixture to either the compaction temperature or
135°C (275°F) for 2 or 4 h. It was reported that WMA was more
sensitive to conditioning temperature than conditioning time. In ad-
dition, the authors indicated that WMA can be more susceptible to
moisture shortly after construction as compared to HMA, but com-
parable performance is expected in the long term.
Newcomb et al. (2015) evaluated the short-term and long-term
laboratory aging protocols for HMA and WMA mixtures. This
study confirmed the outcome of Martin et al. (2014) in that
the effects of plant mixing and processing can be simulated by
2 h of oven aging at either 135°C (275°F) for HMA or 116°C
(240°F) for WMA. Furthermore, it was found that an initial field
performance period of 1 to 2 years (depending on the specific
project climate) can be simulated by 5 days of oven aging at
85°C (185°F), which calls into question the original strategic
© ASCE
J. Mater. Civ. Eng., 2016, 28(11): 04016128
highway research program (SHRP) recommendation that these
conditions simulate 5 to 10 years of field aging. It was also revealed
that while the resilient modulus of WMA was initially lower than
that of HMA for the majority of the field projects, the resilient
modulus of the WMA was comparable to that of the HMA in about
17 months in warmer climates and 30 months in colder climates.
In summary, several research studies have been conducted to
evaluate the aging characteristics of WMA binders and mixtures.
However, most of these studies focused on additive-based WMA
technologies rather than on foamed WMA, which is the most com-
monly used technology in Ohio. This study includes a laboratory
investigation to examine the short-term and long-term aging behav-
ior of laboratory-prepared foamed WMA and HMA mixtures. In
addition, it includes a laboratory investigation to examine the effect
of the extraction and recovery procedures on binder properties.
Objectives of the Study
The primary objectives of this study are to
• Evaluate the short-term and long-term aging characteristics of
foamed WMA as compared to traditional HMA;
• Investigate the effect of the asphalt binder extraction and recov-
ery procedures on the rheological properties of the recovered
binders; and
• Compare the short-term and long-term aging procedures for
asphalt binders and asphalt mixtures.
Testing Program
A laboratory testing program was implemented in this study to ex-
amine the short-term and long-term aging characteristics of foamed
WMA as compared to those of traditional HMA (Fig. 1). As the
figure shows, the laboratory testing program included a binder
aging study and a mixture aging study. The binder aging study
evaluated the short-term and long-term aging characteristics of
the selected asphalt binders. The mixture aging study was divided
into two components. The first component investigated the effect of
the extraction and recovery procedure on the rheological properties
of the two asphalt binders that were used in the preparation of
the laboratory-produced asphalt mixtures. The second component
evaluated the short-term and long-term aging characteristics of
binders recovered from laboratory-produced asphalt mixtures.
The following subsections provide additional details about the
binder and mixture aging studies.
Laboratory Binder Aging
Fig. 2 presents the laboratory testing plan for the asphalt binder
aging. As can be noticed from this figure, two types of asphalt bind-
ers (one polymer-modified PG 70-22 asphalt binder and one neat
PG 64-22 asphalt binder) that are typically used in surface mixtures
in Ohio were included in this study. The short-term aging of the two
asphalt binders was simulated using a Despatch rolling thin film
oven (RTFO) according to AASHTO T 240 (AASHTO 2013a),
and the long-term aging of the asphalt binders was simulated using
a pressure aging vessel (PAV) (Applied Test System, Cheswick,
Pennsylvania) according to AASHTO R 28 (AASHTO 2013c).
A dynamic shear rheometer (DSR) was used to characterize the
viscoelastic behavior of the unaged, RTFO-aged, and PAV-aged
asphalt binders at intermediate and high service temperatures.
Temperature and frequency sweeps were conducted using a re-
search grade DSR device from Rheometric Scientific. The dynamic
shear modulus, G
, and phase angle, δ, were obtained at each
04016128-3 J. Mater. Civ. Eng.
 Laboratory
Testing Program

Binder Mixture
Aging Aging

Effect of Binder Asphalt

Extraction and Recovery Mixture Aging
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Laboratory testing program

Laboratory loading frequency and testing temperature. In addition, Fourier

Binder Aging transform infrared (FTIR) spectroscopy was used to identify and
quantify the amount of functional groups present in the asphalt
binders, and gel permeation chromatography (GPC) was used to
PG 70-22 PG 64-22
determine the molecular size distribution within the unaged,
(A) (B)
RTFO-aged, and PAV-aged asphalt binders. Three replicates were
used for each test.
Physical Chemical Same
Tests Tests as (A)
Effect of Binder Extraction and Recovery
DSR (G*, ) In order to examine the extent of binder aging in an asphalt mixture,
FTIR GPC
Int. + High Temp. it is necessary to extract and recover the asphalt binder from the
mixture. Asphalt binders are generally extracted from mixtures ac-
cording to AASHTO T 164 (AASHTO 2013b) and are recovered in
Unaged Unaged Unaged
accordance with AASHTO T 170 (AASHTO 2009). Because these
procedures introduce a solvent during the extraction and heat during
RTFO RTFO RTFO
the recovery, they are expected to have some effect on the physical
and chemical characteristics of the recovered asphalt binder.
A laboratory testing plan was designed to quantify the effect of
PAV PAV PAV the extraction and recovery procedures on the two asphalt binders
(PG 70-22 and PG 64-22) that were used in the laboratory-
produced asphalt mixtures (Fig. 3). To determine the sensitivity
Fig. 2. Laboratory binder aging
of these asphalt binders to extraction and recovery, controlled

Effect of Binder
Extraction and Recovery

PG 70-22 PG 64-22
(A) (B)

Same
DSR and XRD
as (A)

Original Recovered from Recovered from

Binder Binder + TCE Binder + TCE + Dust

Unaged Unaged Unaged

RTFO RTFO RTFO

PAV PAV PAV

Fig. 3. Effect of binder extraction and recovery

© ASCE 04016128-4 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128

 1000
Laboratory
Mixture Aging

HMA Foamed WMA

at 70oC (kPa)
(A) (B) 10

Recovered
Binders Same as PAV

G*/sin
(A)
0.1 RTFO
DSR, FTIR, and GPC
Unaged
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.

After Mixing Short-Term Long-Term

0.001
0.01 0.1 1 10 100 1000
Fig. 4. Laboratory mixture aging
Radial Frequency (rad/sec)

100000
Binder Air
Tank Tank

at 28oC (kPa)
Foaming 1000
Nozzle

PAV
G* sin
Water 10 RTFO
Tank
Unaged

Control
0.1
Panel 0.01 0.1 1 10 100 1000
Radial Frequency (rad/sec)
Fig. 5. Laboratory-scale asphalt binder foaming device
Fig. 6. Effect of laboratory aging on DSR test results for PG 70-22
binder
amounts of trichloroethylene (TCE), the solvent used in AASHTO
T 164, and dust were added to the unaged, RTFO-aged, and PAV-
aged binders of both PG grades. AASHTO T 164 and AASHTO T 0.3

170 were then used to recover the asphalt binders from the resulting PAV 3
solutions. As can be noticed from Fig. 3, the DSR test was used to
RTFO 2
0.24
characterize the viscoelastic behavior of the original and recovered
Unaged 1
unaged, RTFO-aged, and PAV-aged asphalt binders, and X-ray dif-
fraction (XRD) was used to identify the presence of any limestone
Absorbance

0.18
dust remaining in the recovered asphalt binder. These tests also
aided in determining the effect of the extraction and recovery S=O
C=O
procedures on the PG 70-22 and PG 64-22 asphalt binders that
0.12
were recovered from foamed WMA and HMA mixtures. 3

2
Laboratory Mixture Aging 0.06

Fig. 4 presents the laboratory testing plan for the asphalt mixture 1

aging. As can be noticed from this figure, asphalt mixture aging

0
was evaluated by examining the physical and chemical properties 500 750 1000 1250 1500 1750 2000
of asphalt binders recovered from laboratory-produced foamed Wavenumber (cm-1)
WMA and HMA mixtures at different stages of aging (immediately
after mixing, short-term aging, and long-term aging). AASHTO R Fig. 7. Effect of laboratory aging on FTIR test results for PG 70-22
30 (AASHTO 2013d) was used to simulate the short-term and binder
long-term aging of the asphalt mixtures in the laboratory. To sim-
ulate short-term aging, the aggregates were heated to the mixing
temperature for a minimum of 2 h before being mixed with the was compacted into 150-mm diameter by 100-mm height cylindri-
asphalt binder for approximately 3 min. The loose asphalt mixture cal specimens with a target air void level of 7  0.5% and allowed
was then conditioned in a forced-draft oven for 4 h at 135°C. To to cool to room temperature. The compacted specimens were then
simulate long-term aging, the loose short-term aged asphalt mixture placed in a forced-draft oven for 5 days at 85°C.

© ASCE 04016128-5 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128

 25
Construction and Material Specifications (C&MS) requirements
PAV 3 for Item 442 (Superpave Asphalt Concrete) Type A with a nominal
20 RTFO 2 maximum aggregate size (NMAS) of 12.5 mm. ODOT requires us-
Unaged 1
ing PG 70-22 for Superpave surface mixtures. However, PG 64-22
was included in this study to allow for determining the effect of the
Refractive Index (RI)

15 asphalt binder type on mixture aging.

A Wirtgen WLB10 laboratory-scale asphalt binder foaming
1
device was utilized to foam the asphalt binder by injecting cold
10 water into the heated asphalt binder (Fig. 5). This device consists
2 of an asphalt binder tank, a water tank, an air tank, an asphalt pump,
heating components, a foaming nozzle, air and water pressure reg-
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.

3
5 ulators, and a control panel. It employs a process similar to that
used by large-scale foaming systems that are incorporated into
commercial asphalt plants. A foaming water content of 1.8% by
0 weight of the asphalt binder was used in the production of the
0 4 8 12 16 20
Retention Volume (mL)
foamed asphalt binder. This quantity represents the maximum
water content permitted by ODOT for foamed WMA mixtures.
Fig. 8. Effect of laboratory aging on GPC test results for PG 70-22 In addition, the foamed WMA mixtures were produced at 16.7°C
binder (30°F) lower mixing and compaction temperatures than the
traditional HMA mixtures. This temperature reduction is consistent
with current ODOT specifications for foamed WMA mixtures that
In this study, two asphalt binders (PG 70-22 and PG 64-22) and allow using a compaction temperature 16.7°C (30°F) lower than
one aggregate (limestone) were used in the preparation of the that of the HMA. ODOT, however, does not control the mixing
foamed WMA and HMA asphalt mixtures. The aggregate gradation temperature of the foamed WMA. It is up to the contractor to de-
met the Ohio Department of Transportation (ODOT 2013) termine the appropriate mixing temperature for this material.

14.0
14.0

12.0
G*/sinδδ at 64oC and 10 rad/sec (kPa)

12.0
G*/sinδδ at 70oC and 10 rad/sec (kPa)

11.0
10.0
10.0 9.3 8.7

7.8 8.0
8.0
6.1
6.0
6.0
3.8
3.7 3.8 4.0
4.0 3.3 2.8

1.7 2.0 1.2

1.4
2.0 1.5 1.4 1.0
0.5 0.5
0.0
0.0 Unaged Unaged Unaged RTFO RTFO RTFO PAV PAV PAV
Unaged Unaged Unaged RTFO RTFO RTFO PAV PAV PAV + TCE + TCE + TCE + TCE + TCE + TCE
+ TCE + TCE + TCE + TCE + TCE + TCE + Dust + Dust + Dust
+ Dust + Dust + Dust

5000
5000
G*sinδ at 25oC and 10 rad/sec (kPa)
G*sinδ at 28oC and 10 rad/sec (kPa)

4000
4000

3032
3000
3000
2269
1843
1761 2000
2000 1760

1082 936 941

948 972 833
1000
1000
487 545 272 327
308 176 187

0
0 Unaged Unaged Unaged RTFO RTFO RTFO PAV PAV PAV
Unaged Unaged Unaged RTFO RTFO RTFO PAV PAV PAV
+ TCE + TCE + TCE + TCE + TCE + TCE + TCE + TCE + TCE + TCE + TCE + TCE
+ Dust + Dust + Dust + Dust + Dust + Dust

Fig. 9. Effect of extraction and recovery on PG 70-22 rheological Fig. 10. Effect of extraction and recovery on PG 64-22 rheological
properties properties

© ASCE 04016128-6 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128

 Results and Analysis
Laboratory Aging of Binders
Figs. 6–8 demonstrate the effect of laboratory binder aging on PG
70-22. Similar results were obtained for PG 64-22 asphalt binder.
Therefore, they are not included in this paper. Fig. 6 shows the
effect of aging on G
= sin δ (rutting parameter) obtained at the high
temperature grade and G
sin δ (fatigue parameter) obtained at the
intermediate temperature of PG 70-22 using the DSR test. As the
figure shows, the G
= sin δ and G
sin δ values for the PAV-aged
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.
increase in the larger molecules in the asphalt binder attributable
to aging.
Effect of Extraction and Recovery on Binder Properties
Figs. 9 and 10 present the effect of the extraction and recovery pro-
cedures on the rheological properties of the unaged, RTFO-aged,
and PAV-aged PG 70-22 and PG 64-22 asphalt binders, respec-
tively, as measured using the DSR test. As Fig. 9 shows, little effect
was observed for the unaged and RTFO-aged PG 70-22 asphalt
binders as a result of the extraction and recovery. However, a slight

residue are higher than the RTFO-aged residue, which in turn
are higher than the unaged asphalt binder. The G
= sin δ and
G
sin δ values for the RTFO-aged residue are closer to the unaged
asphalt binder than the PAV-aged residue. Fig. 7 presents the FTIR
spectra for the unaged, RTFO-aged, and PAV-aged PG 70-22 as-
phalt binders. This figure shows an increase in the 1; 700-cm−1
peak corresponding to the carbonyl group (C═O) and the
1; 030-cm−1 peak corresponding to the sulfoxide group (S═O) be-
cause of aging in the RTFO and PAV tests. The increase in the
carbonyl and sulfoxide groups indicates an increase in the number
of large molecules in the asphalt binder, resulting in higher stiffness
and more solidlike behavior. Fig. 8 shows the GPC chromato-
grams for the unaged, RTFO-aged, and PAV-aged PG 70-22 bind-
ers. As expected, the chromatograms for the RTFO-aged and
PAV-aged binders are slightly shifted to the left because of the
decrease in G
= sin δ and G
sin δ was noticed for the PAV-aged PG
70-22 asphalt binder. In addition, by comparing the DSR test re-
sults obtained for the asphalt binders recovered from the binder/
TCE solutions with and without dust, it appears that the extraction
procedure was able to remove most of the dust that was introduced
into the binder/TCE solutions. As mentioned earlier, XRD testing
was performed to examine the presence of any dust remaining in
the recovered unaged, RTFO-aged, and PAV-aged PG 70-22 asphalt
binders. As Fig. 11 shows, the same dominant peaks for the lime-
stone dust were observed in the recovered asphalt binders that were
obtained from the binder/TCE solutions containing dust. This
indicates that some traces of dust remained in the recovered asphalt
binders even though the effect was minimal on the DSR test results.
Fig. 10 shows that the effect of the extraction and recovery
procedures was more pronounced on the rheological properties

15.0
5000 13.0
G*/sinδδ at 70oC and 10 rad/sec (kPa)

Binder Recovered from Unaged+TCE+Dust 1

12.0 11.0
4000 Binder Recovered from RTFO+TCE+Dust 2

Binder Recovered from PAV+TCE+Dust 3

9.0 8.0
3000
Intensity

6.0 5.3
2000 4.4
3.7

1 3.0 2.4
2.1
1.5
1000
3

2 0.0
Unaged HMA WMA RTFO HMA WMA PAV HMA WMA
0 After After STOA STOA LTOA LTOA
0 10 20 30 40 50 60 70 80 90 Mixing Mixing
θ (degrees)
2θ
5000
25000
G*sinδ at 28oC and 10 rad/sec (kPa)

Dust 4000
20000

3000
15000
2269
Intensity

2233
1798
2000
10000 1387 1292
948
1000 667
650
487
5000

0
Unaged HMA WMA RTFO HMA WMA PAV HMA WMA
0 After After STOA STOA LTOA LTOA
0 10 20 30 40 50 60 70 80 90 Mixing Mixing
2θ (degrees)
Fig. 12. Asphalt binder and mixture aging DSR test results for
Fig. 11. XRD test results for recovered PG 70-22 asphalt binders PG 70-22

© ASCE 04016128-7 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128

 of PG 64-22, especially at the intermediate temperature. This was
the case for the unaged, RTFO-aged, and PAV-aged asphalt binders.
This implies that PG 64-22 is more sensitive to the extraction and
recovery procedures using TCE than PG 70-22.
Effect of Mixture Aging on Binder Properties
DSR Test Results
Figs. 12 and 13 present the effect of asphalt binder and mixture
aging on the DSR test results for PG 70-22 and PG 64-22, respec-
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.
the asphalt binder recovered from the LTOA HMA mixture tested
at 70°C. In addition, comparable or slightly higher G
= sin δ and
G
sin δ values were obtained for asphalt binders recovered from
HMA mixtures than those recovered from foamed WMA mixtures.
This was the case for both short-term and long-term aging.
The effect of the extraction and recovery procedures on the
rheological properties of PG 64-22 is obvious in Fig. 13 in that
the G
= sin δ and G
sin δ values for PG 64-22 asphalt binder re-
covered from foamed WMA and HMA asphalt mixtures immedi-
ately after mixing are lower than the unaged PG 64-22 asphalt
binder. In fact, the G
= sin δ values for the asphalt binders recov-

tively. As can be noticed from Fig. 12, the G
= sin δ and G
sin δ
values for PG 70-22 asphalt binder recovered from foamed WMA
and HMA asphalt mixtures immediately after mixing are slightly
higher than the unaged asphalt binder. This indicates that the PG
70-22 asphalt binder undergoes a slight increase in stiffness after
mixing with the aggregates, which can be attributed to the reduced
asphalt binder film thickness and increased exposure to air. It can
also be observed from Fig. 12 that the G
= sin δ and G
sin δ values
for PG 70-22 asphalt binder recovered from short-term oven aged
(STOA), foamed WMA and HMA asphalt mixtures are slightly
higher than the RTFO-aged residue, while the G
= sin δ and
G
sin δ values for PG 70-22 asphalt binder recovered from
long-term oven aged (LTOA), foamed WMA and HMA asphalt
mixtures are slightly lower than the PAV-aged residue except for
ered after mixing are less than 1.0 kPa, the minimum specification
requirement for PG 64-22 at 64°C and 10 rad=s. Furthermore, the
G
= sin δ and G
sin δ values obtained for the PG 64-22 asphalt
binder recovered after mixing in Fig. 13 are close to those obtained
for the PG 64-22 binder recovered from the unaged binder/TCE
solutions with and without dust in Fig. 10.
Similar comparisons between Figs. 10 and 13 show that the
short-term oven aging of PG 64-22 asphalt mixtures results in
higher levels of binder aging than the RTFO procedure, while
the aging caused by the PAV procedure is comparable to that
obtained from long-term oven aging of HMA mixtures and higher
than that obtained from long-term oven aging of foamed WMA
mixtures. Finally, by comparing the G
= sin δ and G
sin δ values

0.04
15.0 0.036
0.035

0.030
G*/sinδδ at 64oC and 10 rad/sec (kPa)

12.0 0.03 0.027

0.026
0.025

8.7
9.0 0.020
0.019
IC=O

0.020
0.02

5.7
6.0

0.01
2.8 2.7
3.0 2.3
1.5
1.2
0.7 0.4
0.00
0.0 Unaged HMA WMA RTFO HMA WMA PAV HMA WMA
Unaged HMA WMA RTFO HMA WMA PAV HMA WMA After After STOA STOA LTOA LTOA
After After STOA STOA LTOA LTOA Mixing Mixing
Mixing Mixing

0.06
5000

0.05 0.046
G*sinδ at 25oC and 10 rad/sec (kPa)

0.045
4000 0.043
0.041 0.041
0.039 0.039
0.04
3032 0.035 0.035
3000
IS=O

0.03

1843
2000
0.02
1183

1000 833 672 701

550 0.01

235
104
0 0.00
Unaged HMA WMA RTFO HMA WMA PAV HMA WMA Unaged HMA WMA RTFO HMA WMA PAV HMA WMA
After After STOA STOA LTOA LTOA After After STOA STOA LTOA LTOA
Mixing Mixing Mixing Mixing

Fig. 13. Asphalt binder and mixture aging DSR test results for Fig. 14. Asphalt binder and mixture aging FTIR test results for
PG 64-22 PG 70-22

© ASCE 04016128-8 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128

 0.04 35%

0.032
27.4%
28% 26.6%
25.4%
0.03 24.5%
0.027 22.8%
0.024 0.024 0.025 0.025
21% 19.3%

LMS (%)
0.021 0.021
IC=O

0.02 0.018 15.4%

14.4%
14%
10.7%

0.01
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.

7%

0.00 0%
Unaged HMA WMA RTFO HMA WMA PAV HMA WMA Unaged HMA WMA RTFO HMA WMA PAV HMA WMA
After After STOA STOA LTOA LTOA After After STOA STOA LTOA LTOA
Mixing Mixing Mixing Mixing

0.06 Fig. 17. Asphalt binder and mixture aging GPC test results for
PG 64-22
0.05 0.045
0.044 0.044
0.039 0.040 0.041 FTIR Test Results
0.04 0.037 The carbonyl and sulfoxide indices suggested by Lamontagne et al.
0.035 0.035
(2001) were utilized to quantify the asphalt binder aging from the
FTIR spectra
IS=O

0.03

Area of the carbonyl band centered around 1,700 cm−1

0.02 IC═O ¼ P
Area of the spectral bands between 2,000 and 500 cm−1
ð1Þ
0.01

0.00 Area of the sulfoxide band centered around 1,030 cm−1

Unaged HMA WMA RTFO HMA WMA PAV HMA WMA
IS═O ¼ P
After After
Mixing Mixing
STOA STOA LTOA LTOA
Area of the spectral bands between 2,000 and 500 cm−1
ð2Þ
Fig. 15. Asphalt binder and mixture aging FTIR test results for
PG 64-22
Spectral normalization was performed by bringing the same
absorbent series at the same point to avoid the variation of binder
film thickness on the KBr plate. This normalization was performed
25%
to compare the test results using the same scale.
Figs. 14 and 15 present the carbonyl and sulfoxide indices ob-
20% 18.1%
tained from the FTIR spectra for PG 70-22 and PG 64-22 asphalt
15.1%
binders, respectively. As expected, the FTIR results show a higher
14.9%
level of aging for PAV-aged binders than for RTFO-aged binders,
15% followed by the unaged asphalt binders. These figures also show
LMS (%)

11.4%
that the increase in the carbonyl indices is more consistent than
10.5% the increase in the sulfoxide indices. Therefore, the change in
10% 7.5% 7.7%
7.1%
the carbonyl index might be a better indicator of the effect of aging.
It can also be observed from these figures that the carbonyl and
sulfoxide indices for asphalt binders recovered from foamed
5% 4.1% WMA mixtures are generally lower than those recovered from
HMA mixtures, which suggests that laboratory-prepared foamed
WMA mixtures undergo slightly lower levels of aging than tradi-
0%
Unaged HMA WMA RTFO HMA WMA PAV HMA WMA tional HMA mixtures.
After After STOA STOA LTOA LTOA
Mixing Mixing
GPC Test Results
Fig. 16. Asphalt binder and mixture aging GPC test results for The GPC data were analyzed by dividing the chromatogram into 13
PG 70-22 slices of equal retention volumes (or elution times) and classifying
the slices into three groups: slices 1–5 for the large molecular size
(LMS), slices 6–9 for the medium molecular size (MMS), and slices
in Fig. 13 for the foamed WMA and HMA mixtures prepared using 10–13 for the small molecular size (SMS). Previous research studies
PG 64-22, it can be noticed that higher levels of binder aging were have reported high correlation between asphalt binder properties
obtained for the HMA mixtures than the foamed WMA mixtures upon aging and the percentage of LMS within the asphalt binder.
for both short-term and long-term aging. Therefore, only the percentage of LMS was used in this study to

© ASCE 04016128-9 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128

 evaluate aging. The percentage of LMS was calculated as the cumu-
lative molecular weight fraction obtained for the first five slices.
Figs. 16 and 17 present the percentage of LMS obtained
from the GPC chromatograms for PG 70-22 and PG 64-22 asphalt
binders, respectively. As these figures show, the percentage of
LMS increased with the increase in the level of aging for both
asphalt binders, with the percentage of LMS being slightly
higher for the traditional HMA mixtures than the foamed WMA
mixtures.
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.
Summary and Conclusions
This study evaluated the short-term and long-term aging character-
istics of two asphalt binders as well as foamed WMA and HMA mix-
tures prepared using these binders. The short-term and long-term
aging of the two asphalt binders were simulated using the RTFO
and PAV tests, respectively, while AASHTO R 30 was used to sim-
ulate the short-term and long-term aging of the laboratory-prepared
asphalt mixtures. The DSR test was used to characterize the viscoelas-
tic properties of the original (unaged, RTFO-aged, and PAV-aged)
and recovered asphalt binders; the FTIR test was used to identify
and quantify the amount of functional groups present in the asphalt
binders; and the GPC test was used to determine the molecular size
distribution within the asphalt binders. In addition, this study inves-
tigated the effect of the extraction and recovery procedures on the re-
covered binder properties. Based on the experimental test results, the
following observations and conclusions were made:
• In general, comparable or slightly higher G
= sin δ and G
sin δ
values were obtained using the DSR test for asphalt binders
recovered from laboratory-prepared HMA mixtures than those
recovered from laboratory-prepared foamed WMA mixtures.
This was the case for both short-term and long-term aged mix-
tures. This indicates that laboratory-prepared foamed WMA
mixtures undergo comparable or slightly lower levels of aging
than traditional HMA mixtures.
• The conventional DSR test results were consistent with the
FTIR and GPC test results in that the carbonyl and sulfoxide
indices from the FTIR and the percentage of large molecular
size from the GPC indicated a slightly higher level of aging
for the laboratory-prepared HMA mixtures than the labora-
tory-prepared foamed WMA mixtures.
• Little effect was observed for the extraction and recovery
procedures on the rheological properties of PG 70-22, especially
for the unaged and RTFO-aged asphalt binders. However, sig-
nificantly lower G
= sin δ and G
sin δ values were obtained for
the recovered PG 64-22 asphalt binder. This implies that the PG
64-22 asphalt binder is more sensitive to the extraction and
recovery procedures using TCE than PG 70-22.
• Little difference in G
= sin δ and G
sin δ was observed for as-
phalt binders recovered from binder/TCE solutions with and
without dust. This suggests that the undertaken extraction pro-
cedure was able to remove most of the dust that was introduced
into the binder/TCE solutions.
• The same dominant peaks for limestone dust were observed
in the XRD test results for recovered asphalt binders obtained
from binder/TCE solutions containing dust. This indicates
that some traces of dust remained in the recovered asphalt
binders even though the effect was minimal on the DSR test
results.
• In general, the G
= sin δ and G
sin δ values obtained for asphalt
binders recovered from short-term oven aged foamed WMA and
HMA mixtures were slightly higher than those obtained for the
corresponding RTFO-aged binders, while the G
= sin δ and
© ASCE 04016128-10
J. Mater. Civ. Eng., 2016, 28(11): 04016128
G
sin δ values obtained for asphalt binders recovered from
long-term oven aged foamed WMA and HMA mixtures were
not consistently higher or lower than those obtained for the cor-
responding PAV-aged binders. This indicates that the RTFO test
procedure results in less aging than the short-term oven aging
procedure specified in AASHTO R 30, while the PAV test pro-
cedure results in comparable aging to the long-term oven aging
procedure specified in AASHTO R 30. Similar results were also
obtained from the FTIR and GPC tests.
Acknowledgments
This study was sponsored by the Ohio Department of Transporta-
tion (ODOT) and the Federal Highway Administration (FHWA).
The authors would like to thank Mr. David Powers and Mr. Eric
Biehl of the ODOT Office of Materials Management for their
valuable contributions to this study.
References
AASHTO. (2009). “Standard method of test for recovery of asphalt binder
from solution by Abson method.” AASHTO T 170-00, Washington, DC.
AASHTO. (2013a). “Standard method of test for effect of heat and air
on a moving film of asphalt (rolling thin-film oven test).” AASHTO
T 240-13, Washington, DC.
AASHTO. (2013b). “Standard method of test for quantitative extraction of
asphalt binder from hot-mix asphalt (HMA).” AASHTO T 164-11,
Washington, DC.
AASHTO. (2013c). “Standard practice for accelerated aging of asphalt
binder using a pressurized aging vessel (PAV).” AASHTO R 28-12,
Washington, DC.
AASHTO. (2013d). “Standard practice for mixture conditioning of hot-mix
asphalt (HMA).” AASHTO R 30-02, Washington, DC.
Ahmed, E. I., Hesp, S. A. M., Samy, S. K. P., Rubab, S. D., and Warburton, G.
(2012). “Effect of warm mix additives and dispersants on asphalt rheo-
logical, aging, and failure properties.” Constr. Build. Mater., 37, 493–498.
Arega, Z., Bhasin, A., Motamed, A., and Turner, F. (2011). “Influence of
warm-mix additives and reduced aging on the rheology of asphalt bind-
ers with different natural wax contents.” J. Mater. Civ. Eng., 10.1061/
(ASCE)MT.1943-5533.0000315, 1453–1459.
Banerjee, A., de Fortier Smit, A., and Prozzi, J. A. (2012). “The effect of
long-term aging on the rheology of warm mix asphalt binders.” Fuel,
97, 603–611.
Gandhi, T., and Amirkhanian, S. (2008). “Laboratory simulation of warm
mix asphalt (WMA) binder aging characteristics, airfield and highway
pavements.” ASCE Conf. Proc., ASCE, Reston, VA, 195–204.
Hasan, M. R. M., Goh, S. W., and You, Z. (2013). “Comparative study on
the properties of WMA mixture using foamed admixture and free water
system.” Constr. Build. Mater., 48, 45–50.
Hossain, Z., and Zaman, M. (2013). “Behavior of selected warm mix
asphalt additive modified binders and prediction of dynamic modulus
of the mixes.” J. Test. Eval., 41(1), 104639.
Kim, H., Lee, S.-J., Amirkhanian, S. N., and Jeong, K.-D. (2013). “Quan-
tification of oxidative aging of polymer-modified asphalt mixes made
with warm mix technologies.” J. Mater. Civ. Eng., 10.1061/(ASCE)MT
.1943-5533.0000479, 1–8.
Lamontagne, J., Dumas, P., Mouillet, V., and Kister, J. (2001). “Compari-
son by Fourier transform infrared (FTIR) spectroscopy of different age-
ing techniques: Application to road bitumens.” Fuel, 80(4), 483–488.
Martin, A. E., et al. (2014). “Evaluation of the moisture susceptibility of
WMA technologies.” Rep. No. 763, National Cooperative Highway
Research Program (NCHRP), Transportation Research Board,
Washington, DC.
Newcomb, D., Martin, A. E., Yin, F., Arambula, E., and Park, E. S. (2015).
“Short-term laboratory conditioning of asphalt mixtures.” Rep. No. 815,
National Cooperative Highway Research Program (NCHRP), Transpor-
tation Research Board, Washington, DC.
J. Mater. Civ. Eng.
 ODOT. (2013). “Construction and material specification.” Ohio Dept. of
Transportation (ODOT), Columbus, OH.
Punith, V. S., Xiao, F., Putman, B., and Amirkhanian, S. N. (2012). “Effects
of long-term aging on moisture sensitivity of foamed WMA mixtures
containing moist aggregates.” Mater. Struct., 45(1), 251–264.
Trujillo, P. (2011). “Long term aging study of WMA binder.” Master’s
thesis, Univ. of Texas at Austin, Austin, TX.
Downloaded from ascelibrary.org by Indian Institute of Technology Roorkee on 05/28/18. Copyright ASCE. For personal use only; all rights reserved.
West, R., et al. (2014). “Field performance of warm mix asphalt technol-
ogies.” Rep. No. 779, National Cooperative Highway Research Program
(NCHRP), Transportation Research Board, Washington, DC.
Xiao, F., Amirkhanian, S. N., and Zhang, R. (2012). “Influence of short-
term aging on rheological characteristics of non-foaming WMA
binders.” J. Perform. Constr. Facil., 10.1061/(ASCE)CF.1943-5509
.0000223, 145–152.

© ASCE 04016128-11 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2016, 28(11): 04016128