Construction and Building Materials 137 (2017) 485–493

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Evaluation of the temperature effect on Rolling Thin Film Oven aging

for polymer modified asphalt
Chuanqi Yan a, Weidong Huang b,⇑, Naipeng Tang a
a
Key Laboratory of Road and Traffic Engineering of Ministry of Education Tongji University, 304 Tongda Building 4800 Cao’an Road, Shanghai 201804, China
b
Key Laboratory of Road and Traffic Engineering of Ministry of Education Tongji University, 608 Tongda Building 4800 Cao’an Road, Shanghai 201804, China

h i g h l i g h t s

A drastic carbonyl jump at certain aging temperature levels was observed for PMA.

Equal aging effect between RTFOT and MRTFOT was detected after carbonyl jump.

Rheology master curves were constructed to validate the findings of FTIR test.

To guarantee fluidity, the aging temperature should be set above 178 °C for PMA.

a r t i c l e i n f o a b s t r a c t

Article history: Rolling Thin Film Oven Test (RTFOT) was found to be inadequate for polymer modified asphalt (PMA)
Received 10 August 2016 since high viscosity PMA will not roll inside the glass bottle during the test. Many suggested that the
Received in revised form 28 January 2017 aging temperature of RTFOT should be promoted for PMA; however the degree of promotion remains
Accepted 29 January 2017
to be decided. To investigate the impact of temperature on RTFOT aging and determine a suitable PMA
Available online 10 February 2017
aging temperature, an innovative approach was proposed by comparing aging effect of RTFOT and
Modified Rolling Thin Film Oven Test (MRTFOT) at different aging temperatures. Infrared spectroscopy
Keywords:
and rheology characteristic were employed to evaluate aging.
Asphalt
PMA
Carbonyl change suggested that with aging temperature lifted, a drastic increase in carbonyl combined
RTFOT with an equal aging effect between MRTFOT and RTFOT occurred at certain aging temperature, and was
MRTFOT inferred to be an indication of PMA obtaining fluidity. PMA rheology master curves exhibited consistent
Aging temperature results with carbonyl change. Based on above findings, a commonly used PMA with 4.5% styrene–buta
Carbonyl diene–styrene should be aged at 178 °C in RTFOT to guarantee fluidity and equal aging effect compared
Infrared spectroscopy with neat asphalt
Rheology master curve Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction facing contractors, suppliers and academia conducted by NCHRP

Rolling Thin Film Oven Test (RTFOT) is one of the most common
used test methods to simulate aging of asphalt plant mixing. It has
been applied for many years and was included in AASHTO and
ASTM standard. However RTFOT has been reported not suitable
for Polymer Modified Asphalt (PMA) [1]. Researches showed that
as a result of the rather lower aging temperature of 163 °C, PMA
with high viscosity will not melt completely and spread into uni-
form thin film [2]. Absence of fluidity in RTFOT leads to insufficient
aging and causes serious problems, one of which was the inappli-
cability of G⁄/sind in evaluating PMA [3]. According to the survey
⇑ Corresponding author.
E-mail addresses: 1992ycq@tongji.edu.cn (C. Yan),
(W. Huang), tangnaixu@126.com (N. Tang).
Project 9-10, PMA short-term aging was considered as one of the
most problematic areas [4].
Due to the drawbacks of RTFOT, researches from two directions
merged in order to find an appropriate way to simulate short-term
aging for PMA.
Some tried to develop whole new aging methods to replace
RTFOT. New equipment such as Modified German Rotating Flask
(MGRF) [5], Stirred Air Flow Test [6] and Rotating Cylinder Aging
Test [7] were proposed in recent years as alternatives to RTFOT
and the latest outcomes were reported in NCHRP Project 09-36
[8]. The report concluded that MGRF was an acceptable alternative
to RTFOT for both neat and modified asphalt binders. However, due
to the inconvenience and financial issue of renewing equipment,
new aging methods like MGRF did not gain popularity [9].
hwd@tongji.edu.cn

http://dx.doi.org/10.1016/j.conbuildmat.2017.01.135
0950-0618/Ó 2017 Elsevier Ltd. All rights reserved.
486 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493
On the other hand, a number of researchers attempted to mod-
ify RTFOT for PMA, so one can age both neat and modified asphalt
without replacing the whole equipment.
The foundation of the RTFOT method is to spread 35 g asphalt
binder into homogeneously thin film with a thickness of 5–
10 lm. Hence the binder will be aged efficiently and uniformly
[10]. Neat asphalt can easily obtain fluidity in aging bottle while
PMA with high viscosity may be too stiff to roll. Thereby, if any
possible measure is taken to force PMA roll in the aging bottle as
neat binder does, the use of RTFOT for PMA will be suitable.
A typically modification is the Modified Rolling Thin Film Oven
Test (MRTFOT) proposed by Bahia [11]. In the MRTFOT, a steel rod
or a number of steel spheres were put into the glass bottle to create
extra shearing forces to force the spreading of thin film. Evalua-
tions indicated that the steel rod was more practical, and was
easier to use and clean after aging. Steel rod of 127 mm long and
by 6.4 mm in diameter was recommended in NCHRP Project 9-10
[4]. However, some investigations reported that the increase in
aging effect was not satisfying and the MRTFOT may not age
PMA fully as expected [12].
Besides the modification in equipment, a number of researches
tended to upgrade the RTFOT by revising the test parameters
[15–17]. And aging temperature and aging period are the most
influential parameters and gain most concerns.
According to AASHTO standard, in the RTFOT, asphalt binder
will be aged in a glass bottle for 85 min at the temperature of
163 °C. The proposer of the RTFOT, California highways, deter-
mined the 85 min test period from actual field calibration on a ser-
ies of asphalt binders. The suitability of this test period has been
confirmed by a correlation study with the calibrated Ottawa sand
mixers [10]. There was rare debate about the 85 min aging period.
Most controversies were raised by the aging temperature. Many
pointed out that the temperature of 163 °C was established on neat
asphalt field investigations and apparently lower than the actual
PMA mixing and planting temperature [1].
Through literature review [17], it was found that there was no
as much as field investigation in the establishment of 163 °C aging
temperature compared with aging period. As the replacement of
the last generation aging method Thin Film Oven Test (TFOT),
RTFOT directly followed the 163 °C aging temperature from TFOT
in 1963. Similarly, TFOT inherited this temperature in 1940 from
the older aging method [18]. The 163 °C aging temperature was
found originating in 1903 when Dow heated asphalt binder and
measured the change in weight and penetration [19]. And back
then, PMA was far from being invented or applied. Hence it is rea-
sonable to doubt the suitability of 163 °C for PMA.
Numerous of researches agreed with increase in aging temper-
ature to enhance the aging effect for PMA. However there was no
clear conclusion about how much aging temperature should be
raised. The level of increase is arbitrary and varies from different
researches in a range of 10–30 °C [15–17]. Some tried to find the
suitable aging temperature for modified asphalt by using
viscosity-temperature curve [20]. It is well-known that based on
a great amount of field investigation, a correlation was found
between mixing temperature and asphalt viscosity, and the opti-
mum neat asphalt viscosity for mixing was 0.17 Pa.s [21]. It seems
reasonable to determine the aging temperature according to the
optimum viscosity. However, the estimated PMA aging tempera-
ture acquired from the optimum viscosity of 0.17 Pa.s often tends
to be higher than expected. At some cases, the estimated temper-
atures can reach and even exceed 200 °C. Thus the reasonability of
0.17 Pa.s regard to PMA is debatable, and the optimum PMA viscos-
ity and aging temperature for the RTFOT remain to be decided.
It should be noticed that rise in temperature improved aging
from two aspects. First, heating-up lowers the viscosity of PMA
and makes binder fluid enough to spread into thin film as expected.
Second, higher temperature leads to more molecular activities in
asphalt and subsequently more chemical reaction with oxygen.
While both sides have significant effects on aging, one cannot tell
which factor governed the enhancement in aging when adjusting
aging temperature, and cannot tell at what temperature PMA can
obtain fluidity in RTFOT.
This study tried to decide whether the PMA in the aging bottle
has obtained fluidity at a certain temperature by comparing the
aging effect between RTFOT and MRTFOT. As mentioned before,
MRTFOT is one of the most famous modifications of RTFOT. And
one important precondition of MRTFOT is the introduction of steel
rod will not affect the aging of neat asphalt. It is because that neat
asphalt has already obtained fluidity at 163 °C without steel rod,
thus extra shearing force has little effect [4]. In other words,
163 °C is suitable for neat binder while PMA requires higher aging
temperature. Based on this assumption, If MRTFOT shows more
serious aging effect than RTFOT at a certain temperature, then
the PMA is still too stiff and the aging temperature should be fur-
ther raised. When the MRTFOT and RTFOT finally show an equal
aging effect, one can consider that at this temperature, PMA has
obtained fluidity and fully aged by RTFOT. Hence, by comparing
the aging effect of RTFOT and MRTFOT, a suitable PMA aging tem-
perature can be determined.
2. Objectives
This study aimed to assess the aging temperature impact on
RTFOT aging regard to PMA, and to propose a method to determine
the suitable RTFOT aging temperature for PMA.
PMAs of different styrene–butadiene–styrene (SBS) content
were aged at different temperatures through RTFOT and MRTFOT
respectively. After aging, infrared spectroscopy and rheology
characteristic of residue were analyzed to evaluate the aging
effect.
It should be noted that the suitable temperature mentioned in
this study is the certain temperature at which PMA obtains fluidity
and can be fully aged by RTFOT. The author holds the opinion that
one should first guarantee the applicability of RTFOT for PMA, and
then it will be appropriate and possible to consider revising the
test parameters from the perspective of practical relevance.
3. Experimental
3.1. Materials
In this study, one neat base binder, one SBS modifier and ele-
mental sulfur were selected to prepare PMA in laboratory. Base
binder was ESSO asphalt (PG64-16). SBS T161B was produced by
DuShanZi Petroleum and Chemical Corporation, China. SBS T161B
was a kind of radial polymer with average molecule weight of
230,000 g/mol, containing 30 wt% of styrene. The amount of SBS
modifier was set at 4.5%, 6.0% and 7.5% by weight of base binder.
In industry production, elemental sulfur and organic sulfide were
commonly used as cross-linking agent to produce storage stable
PMA. In this study, elemental sulfur was used as a cross-linking
agent with 0.15 wt% of base binder.
Following procedure was taken to prepare the SBS modified
asphalt according to the method disclosed in the patent [22].
Firstly, SBS was added to base binder and sheared for 30 min at
180 °C with high shear mixer at the shear speed of 4000 r/min. Sec-
ondly, the blend was stirred for 60 min using mechanical stirrer at
180 °C. Thirdly, cross-linking agent was added to the blend and
stirred for another 240 min at 180 °C. The resulted binders satisfied
the requirements of storage stability test, and because PMAs were
 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493
Table 1
Performance true grades of tested asphalts.
Neat
High temp true grade (°C) 63.3
Low temp true grade (°C) 23.6
produced in large quantities, the aging in production was
minimized.
In this paper, to understand how polymer content affecting
aging effect, PMA of 4.5%, 6.0% and 7.5% SBS content was tested
and named as SBS45, SBS60 and SBS75 respectively. SBS75 repre-
sented the high viscosity PMA used in porous asphalt pavement
which gained popularity in recent years. [23]. Due to a relatively
higher polymer content, the application of RTFOT regarding to high
viscosity PMA might be even more problematic. The performance
true grades of tested asphalts were listed in Table 1.
3.2. Methods
3.2.1. Laboratory aging
In the laboratory short-term aging, aging temperatures were set
to 163, 178 and 193 °C with a 15 °C gap, while the aging period
remained 85 min. All tested asphalts were aged through RTFOT
and MRTFOT, respectively. The 127 mm long, 3.16 mm by diameter
steel rod for MRTFOT was prepared as recommended. Aging bottles
with and without steel rod were presented in Fig. 1.
RTFOT and MRTFOT were both conducted according to AASHTO
T 240. After aging, residues were collected for infrared
spectrometry.
3.2.2. Fourier transform infrared (FTIR) test
To investigate the effect of temperature on aging quantitatively,
an accurate and sensitive index to monitor the aging severity was
expected. According to related studies [24,25], oxygenated func-
tional groups such as carbonyl and sulphoxide are sensitive to oxi-
dation and indexes based on band area ratio of functional groups
are frequently used to quantify aging severity.
In this study, carbonyl instead of sulphoxide was employed due
to its better stability against high temperature [26]. Carbonyl index
was calculated by formula (1), (2) [27]:
X to 75 °C. Based on the sweep tests results, rheological master
A1690 cm1 = A ð1Þ
Fig. 1. Aging bottles with (left) and without (right) steel rod.
487
SBS45 SBS60 SBS75
74.7 79.7 >88
25.9 29.3 33.4
X
A ¼ A1700 cm1 þ A1600 cm1 þ A1460 cm1 þ A1376 cm1
þ A1030 cm1 þ A864 cm1 þ A814 cm1 þ A743 cm1
þ A724 cm1 ð2Þ
As FTIR peak intensities are sensitive to the concentration of
components and the thickness of the sample [28], a reference peak
or several peaks were often used to eliminate the effect of different
concentration and thickness. In this study, the total area of refer-
P
ence peaks is A.
The infrared spectra values of residue were collected using a
Bruker TENSOR FT-IR spectrometer equipped with a reflection dia-
mond ATR accessory. Approximately 1 g asphalt binder was put on
the surface of ATR diamond and fixed by a metal indenter to main-
tain a fully contact between asphalt and ATR diamond. Thirty-two
scans within the wavenumber range of 4000–600 cm1 were
obtained and averaged for each sample. Three replicates of each
sample were scanned. After scanning, computational code devel-
oped in the MATLAB environment was employed to obtain car-
bonyl index automatically to minimize the operative error.
3.2.3. Dynamic shear rheology (DSR) test
Carbonyl content gives quick and precise information on aging,
however it cannot directly reflect the asphalt performance. On the
other hand, rheology characteristic like master curve has been
widely generated and interpreted to evaluate aging for many years.
To validate the findings in FTIR test from the perspective of rheol-
ogy, master curves of complex modulus and phase angle were con-
structed and analyzed.
Frequency sweeps tests between 0.1 and 30 Hz were performed
at temperatures between 5 and 75 °C. The 8 mm diameter, 2 mm
gap, parallel plate testing geometry was used for the tests between
5 and 25 °C and the 25 mm diameter, 1 mm gap was used from 35
curves at a reference temperature of 25 °C were produced. The
modulus master curve was constructed following the sigmoidal
model developed in NCHRP Project A-37A [29], and the phase angle
master curve was modeled based on an effective double-logistic
mathematical function [30].
Asphalts including three SBS contents (0%, 4.5%, and 7.5%) and
four aging conditions (RTFOT at 163 °C, 178 °C, 193 °C and MRTFOT
at 178 °C) were tested.
4. FTIR test results and discussion
Obtained infrared spectrums from wavenumber 1550 to
1850 cm1 were illustrated in Fig. 2 to show the change in
carbonyl.
To analyze aging severity quantitatively, carbonyl indexes were
calculated. And the results of RTFOT aged asphalts were presented
in Fig. 3.
The arrows in Fig. 3 referred to drastic carbonyl increase
between adjacent temperatures. Similar drastic increase was also
found in the result of MRTFOT. In this study, this drastic carbonyl
increase was named as the ‘carbonyl jump’ and was inferred to
be a sign of the obtainment of fluidity. More discussion will be
followed.
488 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493

Carbonyl peak Carbonyl peak

@1690 cm-1 @1690 cm-1

Absorbance
Absorbance

MRTFOT@193°C MRTFOT@193°C

RTFOT@193°C RTFOT@193°C

MRTFOT@178°C MRTFOT@178°C

RTFOT@178°C RTFOT@178°C

MRTFOT@163°C MRTFOT@163°C

RTFOT@163°C RTFOT@163°C

Virgin neat asphalt Virgin SBS45

1850 1800 1750 1700 1650 1600 1550 1850 1800 1750 1700 1650 1600 1550
Wavenumber (cm-1) Wavenumber (cm-1)
(a) (b)

Carbonyl peak
Carbonyl peak @1690 cm-1
@1690 cm-1

Absorbance
Absorbance

MRTFOT@193°C MRTFOT@193°C

RTFOT@193°C RTFOT@193°C

MRTFOT@178°C MRTFOT@178°C

RTFOT@178°C RTFOT@178°C

MRTFOT@163°C MRTFOT@163°C

RTFOT@163°C RTFOT@163°C

Virgin SBS60 Virgin SBS75

1850 1800 1750 1700 1650 1600 1550 1850 1800 1750 1700 1650 1600 1550
Wavenumber (cm-1)
Wavenumber (cm-1)
(c) (d)
Fig. 2. Infrared spectrum of different aging residues: (a) Neat asphalt (b) SBS45 (c) SBS60 (d) SBS75.

250.0
Virgin 163°C Carbonyl 205.6
178°C 193°C jump 194.5
200.0 180.7
Carbonyl index( 10 )
-5

150.0 130.8 129.7

100.6
100.0
67.1
56.2
44.2
50.0
27.5 16.6 6.5
6.2 3.2 5.7
5.8
0.0
0.0 1.5 3.0 4.5 6.0 7.5
SBS content (%)

Fig. 3. Effect of temperature and SBS content on RTFOT aging severity.

4.1. Effect of SBS content SBS may produce more carbonyl, which will also manifest in the
carbonyl index. Consequently, the final effect of increasing SBS
In this study, different SBS content in PMA are supposed to content on aging is determined by the integrated effect of two
affect the aging mainly from two aspects. First, higher SBS content opposite factors.
leads to higher viscosity, which makes it harder for PMA to roll in From Fig. 3, it can be seen that because aging in production was
the aging bottle. Thus inferior aging effect as well as lower car- minimized, virgin asphalts of different SBS content have generally
bonyl index will be detected. Second, according to related studies the same carbonyl index; hence it would be reasonable to quantify
[31,32], at high temperature, SBS is more sensitive to oxygen than aging severity via value of carbonyl index. The result of 163 °C
neat asphalt due to a higher content of carbon-carbon double bond series showed that the increase in SBS content weakened
(C@C). After undergoing the same intensity of aging, asphalt with aging effect.
 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493
It pointed out that under standard RTFOT criteria, the gain in vis-
cosity dominated the integrated effect of increasing SBS content on
aging.
4.2. Effect of temperature
In terms of aging temperature, Fig. 3 exhibited that for the PMA
of a certain SBS content (or the neat asphalt of 0% SBS content), the
higher temperature was lifted, the more serious asphalt aged.
The neat asphalt’s carbonyl growth between temperature levels
was approximately the same, while the growth of PMA apparently
differed among various SBS contents and temperature levels. In
particular, a carbonyl jump was observed for every PMA. Though
each carbonyl jump differed in value, and occurred at different
temperature, they were all the biggest (and much bigger than
the rest) carbonyl increases of its own PMA. Carbonyl jump was
speculated to be the sign of PMA obtaining fluidity and fully aged
by RTFOT.
As mentioned before, heating-up can lower the viscosity and
motivate molecular activity at the same time. The combination of
two effects induced the disproportionate carbonyl growth for
PMA. It was speculated that when the temperature reached a cer-
tain level at which PMA started to flow, the carbonyl jump
occurred. This presumption will be proved in following MRTFOT
result.
The existence of carbonyl jump implied that viscosity of PMA
did not affect aging following simple linear relationship. A certain
temperature may exist above which PMA can rapidly spread into
thin film and be fully aged by RTFOT. And the appearance of car-
bonyl jump was one of the signs that the certain temperature
reached.
4.3. Integrated effect of SBS content and temperature
Since temperature and SBS content can both influence aging by
changing asphalt viscosity, it would be better to consider the inte-
grated effect.
Specifically from Fig. 3, when the temperature was at 163 °C,
carbonyl index went down with SBS content increased. The PMAs
as a whole showed a lower carbonyl index than neat asphalt. After-
wards, with temperature reached 178 °C, a carbonyl jump occurred
to SBS45, while neat asphalt, SBS60 and SBS75 exhibited smaller
growths. It was because that the temperature had reached the cer-
tain point where SBS45 can flow freely and fully aged by RTFOT, yet
SBS60 and SBS75 were still too stiff to roll. Neat asphalt had
already spread into thin film at 163 °C, thus raising temperature
175.0
Neat SBS45
150.0 SBS60 SBS75
Carbonyl index( 10-5)
Carbonyl jump
125.0
102.3
100.0
75.0
56.3
50.5 49.7
50.0 38.0
30.2
24.2
25.0
10.8
0.8
0.0
163-virgin 178-163
Temperature region( )
Fig. 4. Carbonyl growth between adjacent temperature levels in RTFOT.
489
only aggravated aging by motivating molecule activities. It might
be considered that neat asphalt had already experienced carbonyl
jump under 163 °C.
When the temperature was further elevated to 193 °C, carbonyl
jump happened to SBS60 and SBS75 successively. Since SBS45 had
already obtained fluidity and experienced carbonyl jump in 178 °C,
it showed small growth between 178 °C and 193 °C just like neat
asphalt did between 163 °C and 178 °C. At 193 °C, PMAs of all
SBS content had obtained fluidity and were fully aged by RTFOT.
It is interesting that at 193 °C, the relationship between SBS
content and carbonyl index was just the opposite of that of
163 °C. The carbonyl index went up with SBS content increasing,
following a notable linear relationship. The PMAs as a whole
showed a higher carbonyl index than neat asphalt. This was con-
tradictive with the observation of standard RTFOT test; yet consis-
tent with studies reported that SBS modified asphalt showed
higher carbonyl index than neat asphalt after aging [32,33]. This
finding indicated that at a rather high temperature, the effect of
more C@C has dominated the integrated effect of increasing SBS
content, and the aging severity was finally determined by asphalt’s
chemical property (C@C content), not physical property (viscosity).
From this point of view, since neat asphalt had experienced car-
bonyl jump at 163 °C while PMA did not, PMA did not just under-
went insufficient aging compared with field, it actually
experienced less aging than neat asphalt did due to difficulty in
spreading into thin film. The comparison in performance between
PMA residue and neat asphalt residue after RTFOT would be con-
troversial, if they did not undergo the same aging.
To make a better description, after RTFOT, difference between
carbonyl index growths of adjacent temperature levels were sum-
marized in Fig. 4 in terms of temperature and SBS content. For
instance, the value of SBS45 at 178–163 region referred to the
carbonyl growth from 163 °C to 178 °C for SBS45. And the 163 °C
series represented growth from virgin to 163 °C. Three carbonyl
jumps which were above 0.001 (100.0  105)were marked out
in Fig. 4 with orange circle.
From Fig. 4, it is obvious to see the disproportion in PMA’s
carbonyl growth. The distribution of growth was highly related
to temperature region and SBS content. Comparison between
163 °C and virgin asphalt (163-VG) exhibited generally the small-
est growths, while comparison between 193 °C and 178 °C (193–
178) revealed generally the biggest ones. Besides, the higher SBS
content was, the more disproportionate growth was. Taking
SBS75 for example, it showed drastic growth from 178 °C to
193 °C, while barely any change was detected from VG to 163 °C.
This disproportionate distribution of growth regarding to tem-
perature might be misleading. One might think that lifting temper-
ature at a higher level (like 193–178) was more efficient than
lifting at a lower level (like 178–163). However, 193–178 region
held the generally biggest growth because two out of three car-
bonyl jumps occurred in 193–178 region. It was the change in vis-
149.3 cosity induced this disproportion. Actually, in the following
127.4 MRTFOT result it can be seen that due to the extra shearing force
offered by steel rod, all carbonyl jumps were observed in 178–
163 region, making 178–163 region hold the generally biggest
growth (Fig. 7).
There was another interesting fact should be mentioned in
Fig. 4. Neat asphalt shared approximately the same growth
50.9
between each adjacent temperature level with an average growth
of 0.0004151. And after experiencing a carbonyl jump in 178–163,
SBS45 displayed a 0.0005094 growth in 193–178, which was sim-
ilar to neat asphalt’s average growth. Consistent phenomenon was
193-178 observed in the result of MRTFOT.
After experiencing carbonyl jump, both neat asphalt and SBS45
had obtained fluidity. Then lifting temperature only aggravated
aging from the aspect of motivating molecular activity. Thus,
490 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493

250.0
Virgin M-163°C
M-178°C M-193°C
R-163°C R-178°C 192.1 199.6 205.0
200.0 R-193°C Carbonyl jump

Carbonyl index( 10-5)

161.5
152.4 156.0
150.0 136.2

89.8
100.0
70.9

50.0 42.4 40.1

30.9
5.8
6.2 3.2 5.7
0.0
0.0 1.5 3.0 4.5 6.0 7.5
SBS content (%)

Fig. 5. Effect of temperature and SBS content on MRTFOT aging severity.

120.0
Neat SBS45 105.2 By adding steel rod, one can distinguish the effect of lowering
100.0 SBS60 SBS75 viscosity from lifting temperature. Compared with heating-up,
88.9
introducing steel rod will only help asphalt to roll by offering extra
Carbonyl index( 10-5)

80.0 shear force, which is similar to the effect of lowering viscosity.

When the asphalt has obtained fluidity, steel rod will not further
60.0 increase the aging severity. Hence, in this circumstance, RTFOT
43.4
and MRTFOT share the same aging effect. By comparing the aging
40.0
23.5 24.3 effect of RTFOT and MRTFOT, a certain temperature can be deter-
22.7
20.0 11.5 mined at which asphalt obtained fluidity and further lowering vis-
5.5 5.1 cosity will be ineffective.
-1.8 -10.7 -0.6
0.0 Results of MRTFOT aged asphalts were presented in Fig. 5. For
163 178 193 reference, the results of RTFOT were also presented in the figure
-20.0 with dashed line. Only carbonyl index values for MRTFOT were
Temperature region( )
labeled out in Fig. 5. Similar to Fig. 3, the arrows referred to car-
Fig. 6. Aging effect difference between RTFOT and MRTFOT. bonyl jump.
From Fig. 5, it can be detected that carbonyl jump occurred in
MRTFOT as well, but the temperature region at which carbonyl
jump happened advanced by one temperature level. In MRTFOT,
140.0 Neat SBS45
130.6 the steel rod provided extra shear force which partially compen-
SBS60 SBS75
120.0 115.9 sates the lowering viscosity effect of lifting temperature. Hence
Carbonyl jump
Carbonyl index( 10-5)

the PMA could be spread at relatively lower temperatures and car-

100.0
bonyl jump was advanced.
81.6 Fig. 6 summarized the difference in carbonyl content between
80.0
67.6 RTFOT and MRTFOT of all samples tested in this study, which
60.0 may be considered as the effect of steel rod.
47.4 46.4 43.6
39.7 43.5
40.0 36.2 34.3 From Fig. 6, asphalt aged at 178 °C exhibited the biggest differ-
25.2 ences between RTFOT and MRTFOT, followed by 163 °C and 193 °C.
20.0 The aging effect of RTFOT and MRTFOT showed slightly difference
at 193 °C. At 163 °C, PMAs were still far too stiff for steel rod to
0.0
163-virgin 178-163 193-178 force them to roll. And when the temperature reached 193 °C, all
Temperature region( ) PMAs obtained fluidity, extra shear force was no longer useful to
aging. Only at the temperature of 178 °C, when PMAs were at the
Fig. 7. Carbonyl growth between adjacent temperature levels in MRTFOT. edge of flowing freely, the steel rod will force the relatively stiff
PMAs spread into thin film. It can be concluded that the critical
excluding the effect of changing viscosity, heating-up gave neat temperatures at which PMAs started flowing were between
asphalt and SBS45 approximately the same improvement in aging. 178 °C and 193 °C.
From the perspective of SBS content, as expected, neat asphalt
showed no significant difference in aging effect between RTFOT
4.4. Effect of steel rod and MRTFOT at all three temperature levels. For SBS45, due to a
relatively lower viscosity compared with SBS60 or SBS75, steel
Since lifting temperature lowers the viscosity and motivates rod worked effectively for it at 163 °C. But with part asphalt had
molecular activity at the same time. It is difficult to tell which part been rolled into thin film at 163 °C, SBS45 exhibited much smaller
of extra age is induced by lowering viscosity and at what temper- difference at 178 °C compared with SBS60 and SBS75. SBS60 and
ature the effect of lowering viscosity dies out. To solve this prob- SBS75 behaved similarly to SBS45 with great difference at 178 °C
lem, MRTFOT was introduced. and minor difference at 193 °C.
 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493 491

It should be noted that the steel rod provided changeless shear 9

force (only related to its weight and shape) at all temperatures for 8

Log complex modulus (Pa)

all asphalts. However, steel rod-induced carbonyl growth appar-
7
ently depended on SBS content and aging temperature. Since low-
ering viscosity and adding steel rod had a similar effect on 6
spreading films, it proved again that PMA viscosity did not affect 5
aging following simple linear relationship. Virgin neat asphalt
4
For a better description, after MRTFOT, differences between car- RTFOT@163°C
bonyl growths of adjacent temperature levels were summarized in 3 RTFOT@178°C

Fig. 7. Fig. 7 was similar to Fig. 4, and the only difference was that 2 RTFOT@193°C

this was a MRTFOT version. Three carbonyl jumps which were MRTFOT@178°C
1
above 0.0008 (80.0  105) were marked out in Fig. 7.
0
Comparison between Figs. 7 and 4 made it clear to see that car- -5 -4 -3 -2 -1 0 1 2 3 4 5
bonyl jump happened in advance in MRTFOT. Differing from result
Log Reduced Frequency (Hz)
of RTFOT, three carbonyl jumps all occurred in 178–163 region. In
(a)
addition, it was seen that the generally value of carbonyl jump in
MRTFOT was lower than that in RTFOT. As a matter of fact, the 9
growth between adjacent temperature levels distributed more
8
evenly in MRTFOT. This was attributed to the consistent extra

Log complex modulus (Pa)

shear force provided by steel rod at each temperature level. 7
By offering extra shearing force, MRTFOT can force PMA rolling 6
at a relatively lower temperature compared with RTFOT, hence the
5
appearance of carbonyl jump was advanced. Since adding steel rod
Virgin SBS45
is only about rolling PMA, the fact that adding steel rod changed 4
RTFOT@163°C
the occurring temperature of carbonyl jump proved that carbonyl 3 RTFOT@178°C
jump was related to steel rod or viscosity change. RTFOT@193°C
It was also found in Fig. 7 that in 193–178 region, after three 2
MRTFOT@178°C
carbonyl jumps, the growths of all asphalts were around 0.0004 1
with an average of 0.000433. This was consistent with previous
0
findings about SBS45 and neat asphalt. When all asphalts experi- -5 -4 -3 -2 -1 0 1 2 3 4 5
enced carbonyl jump, the growth caused by motivating molecular Log Reduced Frequency (Hz)
activity was rather stable and not sensitive to SBS content or tem- (b)
perature region. In this limited study, the average growth rate was
around 0.0004–0.0005/15 °C. This phenomenon, from the side, 9
showed that motivating molecular activity was not able to induce
intensive carbonyl jump, and carbonyl jump was caused by change 8
Log complex modulus (Pa)

in viscosity. 7
To conclude, carbonyl jump combined with the same aging
6
effect between RTFOT and MRTFOT was the indication of PMA
obtaining fluidity. And by adding steel rod to provide extra shear 5
Virgin SBS75
force, PMAs aged in MRTFOT can experience carbonyl jump at a 4
RTFOT@163°C
lower temperature than RTFOT. Once carbonyl jump happened,
3 RTFOT@178°C
further decrease in viscosity would be ineffective in enhancing
RTFOT@193°C
aging. However, raising temperature can still aggravate aging uni- 2
MRTFOT@178°C
formly by motivating molecular activity. According to the out- 1
comes of FTIR, it can be roughly inferred that in this paper,
0
SBS45 was fully aged at 178 °C and SBS60 and SBS75 were fully -5 -4 -3 -2 -1 0 1 2 3 4 5
aged at 193 °C in RTFOT. Log Reduced Frequency (Hz)
(c)
Fig. 8. Complex modulus master curves: (a) Neat asphalt (b) SBS45 (c) SBS75.
5. DSR test results and discussion

5.1. Modulus master curves
Modulus master curves for neat asphalt, SBS45 and SBS75 were
exhibited in Fig. 8. Different aging procedures including RTFOT at
163 °C, 178 °C, 193 °C and MRTFOT at 178 °C were assessed.
Generally, with the presence of SBS polymer net-work, PMAs
exhibited higher modulus at all frequencies than neat asphalt.
Besides, the introduction of SBS appeared to enhance the aging
resistance of PMA. It can be observed that along with aging, as a
result of asphalt hardening, general increases for complex modulus
were detected for both neat asphalt and PMAs, however the change
occurred to PMAs were less noticeable compared with neat
asphalt, implying less aging impact on modulus.
Specifically to different aging conditions, for neat asphalt,
RTFOT@178 °C and MRTFOT@178 °C revealed overlapped curves,
suggesting a similar aging severity. This phenomenon matched
previous findings derived from carbonyl content that at the aging
temperature of 178 °C, neat asphalt has obtained fluidity. On the
other hand, no major difference was found between RTFOT@178 °C
and RTFOT@193 °C, even though they differed in carbonyl content,
implying that with the increase of oxidation, the impact of oxida-
tion on asphalt rheology properties may gradually decline.
SBS45 showed similar but less significant behaviors with neat
asphalt. RTFOT@178 °C, RTFOT@193 °C and MRTFOT@178 °C
revealed considerably consistent modulus master curves. As dis-
cussed in the infrared spectrometry, SBS45, which contained a rel-
492 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493

90 ification than the complex shear modulus and reduction in phase

80
angle can be considered as an indication of the polymer elastic net-
works [33].
70 Phase angle master curves for neat asphalt, SBS45 and SBS75
Phase angel (degree)

60 were shown in Fig. 9.

The behavior of phase angle was discussed regarding to differ-
50
ent asphalts.
Virgin neat asphalt
40 For neat asphalt, aging induced the transformation from mal-
RTFOT@163°C
30 tene to asphaltene, leading to descending of phase angle master
RTFOT@178°C
curve and a more elastic response. Consistent with that seen for
20 RTFOT@193°C
MRTFOT@178°C
FTIR, RTFOT@178 °C, and MRTFOT@178 °C yielded overlapped
10 master curves. Similar to modulus, RTFOT@193 °C showed minor
0 change compared with that of RTFOT@178 °C, suggesting that the
-5 -4 -3 -2 -1 0 1 2 3 4 5 oxidation impact on asphalt property may gradually decline.
Log Reduced Frequency (Hz) With regard to SBS45, introduction of SBS lowered the phase
(a) angle at all frequencies. But the decrease at low frequencies was
more significant than high frequencies, resulting in a unique
90 reverse bending at low frequencies. The uneven reduction for
80 phase angle at different frequencies was attributed to the rheolog-
ical dominant of SBS phase at low frequencies [34]. As for aging, at
70 high frequencies, SBS45 behaved similarly with those experienced
Phase angel (degree)

60 for neat asphalt, since they all represented the bitumen phase. In
50
the low frequency zone where SBS rheologically dominated, a rise
of the curve towards a more viscous response appeared, which can
Virgin SBS45
40 be attributed to be the degradation of the SBS copolymer [35].
RTFOT@163°C
30 RTFOT@178°C
Identical to that of neat asphalt, curves of RTFOT@178 °C, and
MRTFOT@178 °C overlapped, suggesting that SBS45 has obtained
20 RTFOT@193°C
MRTFOT@178°C fluidity at 178 °C. And comparing RTFOT@193 °C with
10 RTFOT@178 °C, phase angle at high frequencies still yielded no
0 major difference (like neat asphalt) while a marked ascending
-5 -4 -3 -2 -1 0 1 2 3 4 5 was observed at low frequencies, implying the degradation of
Log Reduced Frequency (Hz) SBS occurred with minor aging of bitumen phase.
(b) As for SBS75, the phase angle master curves were further low-
ered. And consistent with modulus, the change caused by aging
90 was weakened by the anti-aging SBS modifier, resulting in a more
80 compacted series of curves compared with neat asphalt or SBS45.
Similar to SBS45, aging caused opposite behaviors between high
70
Phase angel (degree)

and low frequencies. Besides, higher SBS content dominated PMA

60 at an expanded frequency range, making the intersection point
50 where those opposite curves crossed shift towards higher frequen-
40 cies. Specifically, curves of RTFOT@163 °C and virgin SBS75 over-
Virgin SBS75
lapped at high frequencies and deferred at low frequencies,
30 RTFOT@163°C
RTFOT@178°C
suggesting that SBS degradation occurred with ignorable aging of
20 bitumen phase. With temperature lifted to 178 °C, phase angle
RTFOT@193°C
10 MRTFOT@178°C master curve ascended at low frequencies and descended at high
frequencies, implying both SBS and bitumen phase aging aggra-
0
-5 -4 -3 -2 -1 0 1 2 3 4 5 vated. Upon 178 °C, the introduction of steel rod further facilitated
Log Reduced Frequency (Hz) aging at all frequencies, indicating that at 178 °C, SBS75 had not yet
(c) gain fluidity. Compared with MRTFOT@178 °C, RTFOT@193 °C
manifested only minor increase in bitumen phase aging, which
Fig. 9. Phase angle master curves: (a) Neat asphalt (b) SBS45 (c) SBS75. was consistent with result of FTIR test. However when it comes
to low frequencies, a notable growth was observed, suggesting a
atively lower SBS content, behaved similarly to neat asphalt when more severe SBS degradation.
applied with different aging conditions. And at the temperature of According to the DSR test, the changes of modulus and phase
178 °C, it has obtained fluidity and fully aged. Hence the adding of angle at high frequencies were consistent with carbonyl change.
steel rod or further lifting in temperature would not aggravate And those three changes can be considered as indications of bitu-
aging significantly. men phase aging. However, the correlation between carbonyl con-
The modulus master curve of SBS75 further ascended compared tent and rheology characteristics was nonlinear. And with the
with SBS45, due to a higher SBS content. And the changes caused oxidation aggravated, the change in rheology may gradually
by aging were negligible, making further discussion difficult. More decline.
analysis will be presented based on phase angle master curves. On the other hand, the behavior of phase angle at low frequen-
cies represented the aging of SBS phase, which did not follow
5.2. Phase angle master curves exactly the same pattern of bitumen phase aging. At a relatively
high temperature at which PMA has obtained fluidity, further lift-
The phase angle is a measure of the viscoelastic balance of the ing in temperature was not effective in altering bitumen rheology
material behavior. Phase angle is more sensitive to bitumen mod- properties (minor change in modulus or high frequencies phase
 C. Yan et al. / Construction and Building Materials 137 (2017) 485–493
angle), however can still induce notable SBS degradation (increase
in low frequencies phase angle). For PMA, it appeared temperature
not only aggravated aging by lowering viscosity or enhancing
molecular activities, the degradation of polymer was also highly
related to temperature. This, from a side view, indicated that a suit-
able PMA aging temperature was essential.
6. Conclusions and further work
This paper focused on investigating the impact of temperature
on Rolling Thin Film Oven Test (RTFOT) aging effect for Polymer
Modified Asphalt (PMA). The following conclusions can be drawn
from the test results and data analysis:
1. Elevating temperature aggravated aging from two aspects:
reducing viscosity and enhancing molecular activity. Drastic
carbonyl jump was detected when the PMA viscosity was low
enough for PMA to spread into film. After carbonyl jump, fur-
ther lifting in temperature can only aggravate aging uniformly
by facilitating molecular activities.
2. In Modified Rolling Thin Film Oven Test (MRTFOT), as steel rod
can force PMA rolling at a relatively lower temperature than
that of RTFOT. The appearance of carbonyl jump was advanced
by one temperature level (15 °C).
3. After the carbonyl jump was reached, MRTFOT and RTFOT
shared the same aging effect, and PMA finally obtained fluidity
and was fully aged by RTFOT. The appearance of carbonyl jump
and equal aging effect between RTFOT and MRTFOT were two
indications of suitable RTFOT aging temperature for PMA. From
this point of view, it can be roughly inferred that in this paper,
SBS45 was fully aged at 178 °C and SBS60 and SBS75 were fully
aged at 193 °C in RTFOT.
4. Rheology change along with aging was consistent with the car-
bonyl change. Besides, it was observed that polymer degrada-
tion was highly related to aging temperature.
Further work is in progress and tests mentioned in this paper
will be redone with a smaller temperature gap of 5 °C to obtain
more accurate conclusions. Performance tests and viscosity test
will be introduced to further evaluate aging effect. A certain viscos-
ity at which PMA will experience carbonyl jump and be fully aged
will be given.
Acknowledgements
The authors gratefully acknowledge the financial supports by
the National Natural Science Foundation of China under Grant
numbers 51478351.
References
[1] R.B. Mallick, E.R. Brown, An evaluation of superpave binder aging methods, Int.
J. Pavement Eng. 5 (5) (2004) 9–18.
[2] L.I. Ping, Z.Q. Zhang, B.G. Wang, The discussion on the applicability of RTFOT to
modified asphalt, J. Hebei Univ. Technol. (2007).
[3] K.D. Stuart, W.S. Mogawer, P. Romero, Validation of Asphalt Binder and
Mixture Tests That Measure Rutting Susceptibility Using the Accelerated
Loading Facility, Turner-Fairbank Highway Research Center, McLean, VA, 2000.
[4] H.U. Bahia, D.I. Hanson, M. Zeng, Characterization of Modified Asphalt Binders
in Superpave Mix Design, 2001.
493
[5] S. Ramaiah, J. D’Angelo, R. Dongré, Evaluation of modified german rotating
flask, Transp. Res. Rec. 1 (2004) 80–88.
[6] C.J. Glover, R.R. Davison, N. Vassiliev, A New method for simulating hot-mix
plant asphalt aging, Binder Content (2002).
[7] A.F. Verhasselt, Chapter 17 A kinetic approach to the aging of bitumens, Dev.
Pet. Sci. 40 (2000) 475–497.
[8] D.A. Anderson, R. Bonaquist, Investigation of short-term laboratory aging of
neat and modified asphalt binders, NCHRP Rep. (2012).
[9] J. Jia, X.N. Zhang, Y. Yuan, Rolling thin film oven test investigation for polymer
modified asphalt, J. Harbin Inst. Technol. 12 (6) (2005) 635–638.
[10] F.N. Hveem, E. Zube, J. Skog, Proposed new tests and specifications for paving
grade asphalts, Assoc. Asphalt Paving Technol. Proc. (1963).
[11] H. Bahia, H. Zhai, A. Rangel, Evaluation of stability, nature of modifier, and
short-term aging of modified binders using new tests: LAST, PAT, and modified
RTFO, Transp. Res. Rec. 1638 (1998) 64–71.
[12] M. Hemsley, An evaluation of the RTFO modified test method, Report prepared
for the Southeastern Asphalt User-Producer Group (SEAUPG), Ergon, PO Box,
1999, 1639: 39215-1639.
[13] R. Satish, Evaluation of the Modified Rolling Thin Film Oven Test Method,
Report prepared for the TRB Asphalt Binder Expert Task Group, Federal
Highway Administration, Washington, DC, 2000.
[14] J. Jia, X.N. Zhang, Modification of the rolling thin film oven test for modified
asphalt, in: SATC, 2005.
[15] J. David, A. Zia, Development of rubber binder specifications in California:
project update, Asphalt Binder Expert Task Group Meeting, 2015.
[16] B. Hofko, M. Hospodka, Impact of RTFOT and PAV conditioning parameters on
viscoelastic behavior and binder performance grade, Transportation Research
Board 95th Annual Meeting, 2016 (16-1383).
[17] C.A. Bell, Summary report on aging of asphalt-aggregate systems, Asphalt
Mixtures (1989).
[18] R.H. Lewis, J.Y. Welborn, The properties of the residues of 50–60 and 85–100
penetration asphalts from oven tests and exposure, Public Roads (1941).
[19] A.W. Dow, Asphalt experiments at Washington, Eng. Rec. 47 (18) (1903).
[20] A. Simone, M. Pettinari, F. Petretto, et al., The influence of the binder viscosity
on the laboratory short term aging, Procedia – Soc. Behav. Sci. 53 (2290) (2012)
421–431.
[21] Y. Yildirim, Mixing and compaction temperatures for Superpave mixes,
Pratique Médicale & Chirurgicale De Lanimal De Compagnie 40 (2000) 137–
140.
[22] P. Maldonado, J. Mas, T.K. Phung, Styrene block polymers, US4145322, 1979-
03-20.
[23] P.A. Morgan, R.E. Stait, S. Reeves, et al., The feasibility of using twin-layer
porous asphalt surfaces on England’s strategic road network, Trl Published
Project Report, 2010.
[24] C. Yan, W. Huang, Q. Lv, Study on bond properties between RAP aggregates and
virgin asphalt using Binder Bond Strength test and Fourier Transform Infrared
spectroscopy, Constr. Build. Mater. 124 (2016) 1–10.
[25] G. Dondi, F. Mazzotta, A. Simone, et al., Evaluation of different short term aging
procedures with neat, warm and modified binders, Constr. Build. Mater. 106
(2016) 282–289.
[26] J. Claine Petersen, Asphalt oxidation - an overview including a new model for
oxidation proposing that physicochemical factors dominate the oxidation
kinetics, Fuel Sci. Technol. Int. 11 (1) (1993).
[27] V. Mouillet, J. Lamontagne, F. Durrieu, et al., Infrared microscopy investigation
of oxidation and phase evolution in bitumen modified with polymers, Fuel 87
(7) (2008) 1270–1280.
[28] J. Lamontagne, P. Dumas, V. Mouillet, et al., Comparison by Fourier transform
infrared (FTIR) spectroscopy of different aging techniques: application to road
bitumens, Fuel 80 (4) (2001) 483–488.
[29] N.I.M. Yusoff, M.T. Shaw, G.D. Airey, Modelling the linear viscoelastic
rheological properties of bituminous binders, Constr. Build. Mater. 25 (5)
(2011) 2171–2189.
[30] S.M. Asgharzadeh, N. Tabatabaee, K. Naderi, et al., An empirical model for
modified bituminous binder master curves, Mater. Struct. 46 (9) (2013) 1459–
1471.
[31] X. Lu, U. Isacsson, Chemical and rheological evaluation of ageing properties of
SBS polymer modified bitumens, Fuel 77 (9) (1998) 961–972.
[32] Lu. Xiaohu, Ulf. Isacsson, Artificial aging of polymer modified bitumens, J. Appl.
Polym. Sci. 76 (12) (2000) 1811–1824.
[33] M.L. Huggins, Viscoelastic properties of polymers, J. Am. Chem. Soc. (19)
(1961) 4110–4111.
[34] G.D. Airey, Rheological properties of styrene butadiene styrene polymer
modified road bitumens, Fuel 82 (14) (2003) 1709–1719.
[35] S. Linde, U. Johansson, Thermo-oxidative degradation of polymer modified
bitumen, in: Polymer Modified Asphalt Binders, Symposium.