Available online at www.sciencedirect.

com

ScienceDirect
International Journal of Pavement Research and Technology 10 (2017) 322–332
www.elsevier.com/locate/IJPRT

Investigation on performances of asphalt mixtures made with

Reclaimed Asphalt Pavement: Effects of interaction between virgin
and RAP bitumen
Luca Noferini a,b,⇑, Andrea Simone a, Cesare Sangiorgi a, Francesco Mazzotta a
a
University of Bologna, DICAM Department of Civil, Chemical, Environmental and Materials Engineering, Viale Risorgimento 2, 40136 Bologna, Italy
b
GHD Pty Ltd, Geotechnics and Pavement Engineering, 999 Hay Street, Perth, WA, 6000, Australia

Received 27 May 2016; received in revised form 16 November 2016; accepted 29 March 2017
Available online 25 April 2017

Abstract

According to most recent surveys, the European area produced 265 mil tonnes of asphalt for road applications in 2014. In the same
year, the amount of available RAP was more than 50 mil tonnes. The use of RAP in new blended mixes reduces the need of neat bitumen,
making RAP recycling economically attractive. Despite the economic and environmental benefits, road authorities tend to limit the use
of RAP in asphalt mixes due to uncertainty about field performances. The present study focuses on the interaction between neat and
RAP bitumen in asphalt mixes made with different RAP content. The effects of RAP on physical and rheological properties of the final
bituminous blend were investigated. This study is part of a wider research, where a specific type of asphalt mixture was produced with
different RAP contents being 10%, 20% and 30% by mass of the mix. Bitumen was extracted and recovered from asphalt mixes, then it
was subjected to the following laboratory tests: standard characterization, dynamic viscosity and rheological analysis with DSR. Find-
ings showed that the effects of RAP bitumen on the final blend varied in proportion to RAP content. A threshold value of RAP content
was found, below which bitumen was not subjected to significant changes in physical and rheological properties. Practical implications
on production methods and paving of RAP mixes are also proposed.
Ó 2017 Chinese Society of Pavement Engineering. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Reclaimed Asphalt Pavement (RAP); Recycling; Bitumen blending; Bitumen rheology

1. Introduction the required new bitumen content, makes the use of RAP
in HMA mixtures economically attractive [2]. It is consid-
Recycling hot mix asphalt results in a reusable mixture ered that the most economical use of RAP is in the interme-
of aggregates and aged asphalt binder known as Reclaimed diate and surface layers of flexible pavements because
Asphalt Pavement (RAP) [1]. Using the old asphalt bitu- the less expensive RAP binder can replace a portion of
men in the newly blended mixtures and, therefore, reducing the more expensive virgin binder [3]. Hundred per cent of
the reclaimed asphalt can be recycled [4] with different
methods: hot recycling in asphalt plant, hot in-place recy-
⇑ Corresponding author.
cling, cold in-place recycling and full depth reclamation
E-mail addresses: luca.noferini2@unibo.it, luca.noferini@ghd.com
(L. Noferini). are the most commonly applied techniques [5].
Peer review under responsibility of Chinese Society of Pavement The percentage of RAP (usually expressed by mass of
Engineering. the mix) that can be incorporated into asphalt mixtures

http://dx.doi.org/10.1016/j.ijprt.2017.03.011
1996-6814/Ó 2017 Chinese Society of Pavement Engineering. Production and hosting by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332 323

depends on production process (plant type, production 2. Experimental plan

temperature, mixing time, and discharge temperature),
paving technology and permitted emissions as well as 2.1. Objectives and research approach
RAP source and properties. For instance, RAP obtained
from high-trafficked roads (highways or freeways) is likely The study aims at investigating the effects of the
to contain polymer modified bitumen, which limits bitu- interaction between virgin and RAP binder on physical
men oxidation with beneficial effects on the RAP mix and rheological properties of the composite bituminous
[6]. These factors affect the interaction between RAP blend. The objectives of the research are summarized as
and virgin bitumens and consequently impact the perfor- follows:
mance of asphalt mixtures [7]. The maximum amount of
reclaimed asphalt is mainly limited by the production
Investigate the effects of the presence of RAP into
technology [7]. Considering hot in plant recycling, most asphalt mixtures on standard characteristics of bitumen
conventional drum plants can accommodate 50% RAP, as penetration grade and softening point.
whereas the percentage of reusable RAP in batch plant
Determine the effects of RAP binder on rheology of
ranges from 10 to 30% [8]. These limitations have been composite blends with neat binder.
overcome thanks to multiple technologies readily avail-
Define threshold values for RAP content above which
able for production of up to 100% recycled hot mix dynamic characterization tests are needed, in addition
asphalt [9]. to standard characterization tests, in order to evaluate
In Italy, RAP availability is 10 million tonnes, but the effects of the interaction between neat and RAP
only 20% of the recovered material is used in hot/warm binder.
recycling processes [10]. This means that RAP is accumu-
Identify practical implication on the use of different
lating in stockpiles, used in low-value [11] and non- RAP percentages into asphalt mixtures.
bituminous applications (e.g. aggregate in unbound
layers), or being dumped. In Germany, France, and The present study is part of a wider research on perfor-
The Netherlands, the available quantity of RAP is mances and durability of asphalt mixtures made with RAP
11.5, 6.9 and 4.5 million tonnes respectively. The percent [13]. The research is divided in two phases: in the first
of available RAP used in hot/warm recycling is 90% in phase, the effects of RAP on asphalt mixtures are investi-
Germany, 64% in France, and 76% in The Netherlands. gated. A specific typology of asphalt mixture is produced
The European average is 58%. The European Union with different RAP percentages of 10%, 20% and 30% by
released guidelines to favour recycling and re-using of weight of the mix. Asphalt materials are characterized
waste or by-products, including RAP. According to and compared in terms of resistance to fatigue, stiffness
Directive 2008/98/EC, Member States of EU shall take modulus and volumetric properties. The second phase,
measures to promote the re-use of products and prepar- which is reported in this paper, aims at investigating the
ing for re-use activities [12]. The new circular economy effects of RAP on asphalt binder as a composite blend of
package issued by the European Commission in Decem- neat and RAP bitumen. Bituminous samples are recovered
ber 2015 aims at increasing the use of recycled material from asphalt samples tested in phase 1 and further charac-
and RAP in road construction by promoting cradle-to- terized and compared with standard and dynamic tests.
cradle strategy to save raw materials, reduce carbon The objective of the research is to comprehensively char-
footprint of construction process and save money. Nev- acterize RAP mixtures in terms of performances of asphalt
ertheless, re-use and hot/warm asphalt recycling practices mixtures and bitumens.
struggle in Italy.
Many agencies and public administrations insert
restrictions on RAP percentages ranging from 10 to 2.2. Materials and methods
30% in their regulations due to concerns for pavement
performances and production technologies. Uncertainties Five bitumens were investigated. All the bitumens were
concern the interaction between RAP binder and virgin recovered from specimen of asphalt mixture tested to fati-
binder. Inaccurate assumptions on the effects of interac- gue in the previous phase of the research. The asphalt mix-
tion might create problems both in mix design and pave- tures were produced incorporating different percentages of
ment performance [5,7], leading to mixtures that might be RAP into a traditional asphalt concrete, which was used as
subjected to premature failure to cracking, ravelling, reference mixture. Asphalt mixes are classified as Mix0
moisture damage and rutting. At present, there is no (control mixture), Mix1 (control + 10% RAP), Mix2 (con-
industrial approved standard method to predict the degree trol + 20% RAP) and Mix3 (control + 30% RAP). A sam-
of blending in laboratory [3], and tests shall be carried out ple of bitumen was recovered from each mixture,
for evaluating the effects of interaction between RAP bin- respectively B0 from Mix0, B1 from Mix1, B2 from Mix2,
der and virgin binder on mechanical behaviour of RAP B3 from Mix3 and BRAP from the unprocessed RAP. Bitu-
mixtures. minous samples B1, B2 and B3 are considered being blends
324 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332

Table 1
Composition of asphalt mixtures, percentages by mass of aggregates.
MATERIAL MIX0 (control) MIX1 (10%RAP) MIX2 (20%RAP) MIX3 (30%RAP)
Gravel 10/20 mm 25 26 26 23
Gravel 6/10 mm 16 13 14 15
Gravel 4/6 mm 16 16 14 15
Sand 0/4 mm 40 33 24 15
Filler 3 2 2 2
RAP 0/8 mm 0 5 8 14
RAP 8/12 mm 0 5 12 16
Recovered bitumen from RAP 0/8 0.0 0.3 0.4 0.7
Recovered bitumen from RAP 8/12 mm 0.0 0.3 0.6 0.8
Virgin bitumen Pen Grade 50/70 5.0 4.5 4.0 3.5

because of the presence of both neat and RAP binder. The incremental temperature rate of 0.5 °C/min [15]. The test
composition of the asphalt mixtures is given in Table 1. procedure with 8 mm plane plate and 2 mm gap was stud-
Asphalt mixtures were manufactured in laboratory with ied. In the TS test the complex shear modulus (G*) and the
design binder content of 5% by mass of aggregates and phase angle (d) were measured with DSR and studied in
design air voids content of 5% by volume of the mix. terms of master curves.
Aggregates were heated at 180 °C and RAP was dried at
110 °C for 2 h before mixing. Neat binder was incorporated 3. Results
into mixes considering the presence of aged binder into
RAP fractions. Significant effort was put forth to achieve 3.1. Standard characterization test results
similar grading, air voids and binder content in each spec-
imen, ensuring that results of laboratory tests are mainly Results in Table 3 show a decrease in penetration and an
dependent on the percentage of RAP included. increase in softening point as RAP content increases. This
Asphalt binder was recovered from prismatic specimens indicates a hardening of the composite blend caused by the
and unprocessed RAP using a centrifuge extractor accord- presence of aged bitumen. Penetration and softening point
ing to the European standard EN 12697-1 annex B.1.5 for change in proportion to the RAP content in the mixture.
cold extractions methods, using dichloromethane CH2Cl2 Parameters measured for B0, B1 and B2 indicate a small
organic solvent. The solution of bitumen and DCM from hardening of the composite blend due to the presence of
prior extraction was distilled with rotary evaporator, RAP binder. The hardening effect of the old bitumen on
according to EN 12697-3. The extracted binder content is the composite blend becomes relevant as the RAP binder
reported in Table 2. content is above 20% by weight of the mix.
Recovered binders were tested for determining penetra- The Penetration Index (PI) was calculated according to
tion grade (EN 1426), softening point (EN 1427) and EN 12591 as an indicator of temperature susceptibility of
dynamic viscosity at 160 °C (EN 13702-2). Empirically the bitumen. The lower the value of PI, the higher the tem-
based tests, such as penetration, softening point as well perature susceptibility of the binder. PI reported in Table 3
as the more fundamental viscosity, are able to describe indicates that B0, B1 and B2 have identical temperature sus-
the changes in rheological performance of neat and aged ceptibility, with B3 and BRAP having a higher PI value and,
bitumen blends [14]. therefore, a lower temperature susceptibility.
Rotational Viscometer (RV) was used to evaluate the
binder viscosity at high temperatures. The RV measures 3.2. Dynamic viscosity test results
the torque required to rotate a spindle at constant speed
while immersed in the simple fluid. Dynamic viscosity is Viscosity values of bituminous blends for each test tem-
proportional to this measured torque. For all bitumen perature are shown in Table 4.
the dynamic viscosity (g) was calculated at the temperature
of 135, 150, 160 and 170 °C. In order to validate the test,
the sample was thermo-regulated at 5 °C at intervals of
Table 2
7 min and 10 °C at intervals of 12 min. Tests were con- Soluble binder content extracted and recovered from beams and unpro-
ducted according to EN 13702-2. Viscosity can be consid- cessed RAP.
ered to be a measurement of purely Newtonian flow and Specimen Recovered binder Binder content bc
therefore independent of strain rate at test temperatures sample code (% on mixture mass)
[14]. Mix0 B0 5.13
The DSR was used to measure the linear viscoelastic Mix1 B1 4.92
properties of bitumens using a sinusoidal loading mode. Mix2 B2 5.01
A temperature sweep (TS) was applied over the range from Mix3 B3 5.07
0 °C to 90 °C, at a fixed frequency of 10 rad/s with an Unprocessed RAP BRAP 4.87
 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332 325

Table 3 ified Binders (PMB), for which production and working

Results of standard characterization tests on recovered binders. temperatures range from 160 to 180 °C. However, the pres-
Bituminous Penetration at 25 °C Softening point Penetration ence of polymers significantly differentiates the rheological
sample (dmm) EN 1426 (°C) EN 1427 Index EN behaviour of modified and unmodified binders, allowing
12591
the firsts to be subjected to higher temperatures without
B0 27 55 1.3 experiencing ageing. Lowering the viscosity of RAP blends
B1 25 56 1.3
B2 21 57 1.3
appears to be crucial to ensure performances and feasibility
B3 16 63 0.8 of RAP asphalt mixtures. An increase in production and
BRAP 7 73 0.5 working temperatures might cause the composite blend to
harden, leading to a further increase in viscosity. There-
fore, a softer neat bitumen or additives may be used for
lowering the viscosity of RAP blends.
Table 4
Results of dynamic viscosity test on recovered binders. Results indicate no significant differences in viscosity
between B0 and B1. Therefore, the presence of 10% or less
Bituminous sample Dynamic viscosity (Pas)
of RAP (low RAP content [17]) into the mixture does not
135 °C 150 °C 160 °C 170 °C
affect the viscosity of the composite blend. Viscosity signif-
B0 1.36 0.46 0.32 0.25 icantly increases as RAP content exceeds 20% (high RAP
B1 1.41 0.45 0.33 0.25
content [17]) at temperatures of 150 and 160 °C. This indi-
B2 1.51 0.46 0.37 0.28
B3 1.59 0.75 0.53 0.32 cates a significant hardening effect on the blend caused by
BRAP 1.92 0.96 0.55 0.32 the aged binder as the RAP content is above 20%.
The temperature susceptibility of bituminous blends was
represented investigating the trend of viscosity as function
Viscosity of blends increases in proportion to RAP bin- of temperature (Fig. 2). Because viscosity measurements
der content (Fig. 1). The higher the RAP binder, the higher are fitted with power-law trend, the x-coefficient of
the viscosity of the blend. The presence of the more viscous power-law can be used as indicator of temperature suscep-
old binder causes the composite blend to harden, resulting tibility of bitumens. Results show that the higher the RAP
in an increase in viscosity of the mixture. This confirms the binder content, the higher the temperature susceptibility
results obtained from penetration and softening point tests. within the test temperature range. The trend of viscosity
RAP binder has been exposed to ageing processes that of RAP blends is between the trends of neat and RAP bin-
have hardened the bitumen. Short term (ST) ageing, which ders in proportion to RAP binder content in the blend.
takes place during mixing, transport and laying of bitumi- When the RAP content is above 20%, the bituminous blend
nous materials, changes the chemical composition of the experiences a significant shift towards higher values of vis-
binder with significant impact on its rheological properties. cosity. This indicates a hardening effect of RAP binder pre-
During ST ageing, bitumen experiences increasing in vis- sent in the mixture.
cosity caused by oxidation, volatilisation and exudation Viscosity test results can be used to verify the assump-
([5,16]. During the long term (LT) ageing, bitumen progres- tion of total blending between neat and RAP binders after
sively increases viscosity and stiffness due to oxidation, being extracted with the procedure previously described.
polymerisation, photo-oxidation of surface layers, thixo- The procedure of extraction and recovery of bituminous
tropy and syneresis [1]. The major part of asphalt ageing blends is expected to produce 100% blending of RAP and
occurs during the short term [17]. The increase in viscosity virgin binder [5]. The assumption of total blending was ver-
of RAP bitumen is caused by several factors: ratio of resins ified using the log-additivity rule (LAR). LAR predicts vis-
to asphaltenes [14], increase in number of asphaltenes cosity of homologous polymer blends (HPB), a sub-class of
(between 5 to 20% by weight [18]), increase in molecular PB with chemically identical polymers differing in molar
weight and polydispersity [16]. Therefore, the presence of mass. Viscosity of HPB proportionally depends on mass
RAP binder changes chemistry of the bituminous blend, fraction (percentage by mass) of constituent polymers
resulting in a harder and consequently more viscous com- [19]. LAR is expressed as [20]:
posite bitumen. X
The values of viscosity of RAP blends are greater than ln gC ðT Þ ¼ wi ln gi ðT Þ
i
typical viscosities of 50/70 bitumens. This is expected since
penetration grade and softening point of the neat binder In which gC is the viscosity of the composite blend at
showed deviation from typical values of unmodified temperature T, wi and gi are the mass fraction and the vis-
50/70 binders (Table 3), indicating a harder bitumen com- cosity of the component i at temperature T. Results in
pared with a traditional 50/70. Measured viscosities exceed Table 4 show that viscosity of bitumen blends increases
recommended values of 0.17 ± 0.02 and 0.23 ± 0.03 Pas at proportionally with the aged binder content in the compos-
160 °C reported in ASTM D 2493 respectively for optimal ite blend, ranging from recovered neat binder (B0) viscosity
mixing and compaction of HMA with unmodified binders. and RAP binder (BRAP) viscosity. Similarly to HPB, the
Similar values of viscosity may be found on Polymer Mod- LAR was applied to bitumen blends. Viscosity was calcu-
326 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332

Fig. 1. Dynamic viscosity as function of aged binder content.

2.5

2.0
B0
Dynamic viscosity (Pa∙s)

B1
1.5 B2
B3
BRAP
1.0 B0 (y=1.25x^-1.23)
B1 (y=1.28x^-1.24)
B2 (y=1.35x^-1.20)
0.5 B3 (y=1.62x^-1.11)
BRAP (y=2.06x^-1.26)

0.0
130 140 150 160 170 180
Test temperature (°C)

Fig. 2. Temperature susceptibility of bituminous samples measured through viscosity tests.

lated with LAR for B1, B2 and B3, then predicted and mea- considered as homologous blends, validating the assump-
sured values were compared. Results are reported in tion of total blending between neat and RAP binders for
Table 5. the recovered bitumens.
A strong correlation was found between measured and
calculated viscosities, as indicated by the linear relationship 3.3. Dynamic mechanical analysis with dynamic shear
between ln(gM) and ln(gC) with slope of 0.99 and R2 = 0.99 rheometer
(Fig. 3). The viscosity of bitumen blend is proportionally
affected by the content of RAP binder in the composite 3.3.1. Complex modulus isochronal plots (master curves)
mix, showing a viscosity-concentration dependence in rela- An isochronal plot is a curve on a graph representing the
tion to the LAR. Investigated bituminous blends can be behaviour of a system at a constant frequency or loading
 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332 327

Table 5
Comparison between viscosity values of bituminous blends measured in laboratory and values calculated with LAR (viscosity is expressed in natural log
values).
Bituminous Dynamic viscosity at 135 °C Dynamic viscosity at 150 °C Dynamic viscosity at 160 °C Dynamic viscosity at 170 °C
sample
Measured Calculated Measured Calculated Measured Calculated Measured Calculated
B0 0.307 0.307 0.777 0.777 1.139 1.139 1.386 1.386
B1 0.344 0.348 0.799 0.799 1.109 1.109 1.386 1.386
B2 0.412 0.387 0.777 0.777 0.994 0.994 1.273 1.273
B3 0.464 0.424 0.288 0.288 0.635 0.635 1.139 1.139
BRAP 0.652 0.652 0.041 0.041 0.598 0.598 1.139 1.139

Fig. 3. Relationship between measured and LAR calculated viscosity of bituminous samples.

time. Curves of the complex modulus G* versus tempera- temperature cracking susceptibility of the blends. As tem-
ture are isochrones [21]. perature exceeds 20 °C, isochrones of B1, B2 and B3 deviate
Master curves of complex modulus G* are reported in from the reference curve of B0 and shift towards higher val-
Fig. 4. Values of complex modulus (Table 6) increase in ues of G* (with a magnitude in proportion to the RAP bin-
proportion to RAP content over the test temperature der content). The stiffening effect of the RAP binder on the
range, resulting in a vertical shift of G* curves. The complex modulus of the blends becomes more relevant
increase in complex modulus indicates a hardening of the when temperature exceed 20 °C, becoming constant at a
bitumen caused by the presence of the aged binder. temperature of approximately 50 °C.
The hardening effect of aged binder on the composite
blend is proportional to the RAP binder content in the 3.3.2. Phase angle isochronal plots
mixture. When RAP content is equal or below 10%, the The phase angle d is plotted in Fig. 5 as tgd versus
vertical shift is almost negligible. For RAP percentage temperature. Values of d are reported in Table 7. Tgd
equal or above 20%, the influence of recovered binder is defined as the ratio between loss (G00 ) and storage
becomes more relevant, resulting in a significant vertical (G0 ) components of the complex modulus, therefore it
shift of G* curves. Therefore, adding small amounts of can be considered as an indicator of the viscoelastic bal-
reclaimed asphalt into the mixture has very limited effects ance of the mixture.
on binder stiffness. Results show that the phase angle decreases in propor-
The effects of the aged binder on the complex modulus tion to RAP content in the mixture over the temperature
vary over the test temperature range. At temperatures from range. Since tgd is defined as G00 /G0 , the decrease in the
0 °C to 20 °C, curves of G* for B0, B1, B2 and B3 show the phase angle corresponds to an increase in the elastic
same trend and similar values, indicating a limited stiffen- behaviour of the bitumen caused by the presence of aged
ing effect of the aged binder. This might suggest that the binder.
presence of even high percentages (greater than 20%) of As observed for complex modulus isochrones, the mag-
RAP binder into the mixtures has small effect on the low- nitude of the decrease in tgd and hence the increased elastic
328 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332

1.E+09

1.E+08

1.E+07
Complex modulus G* (Pa)
1.E+06

B0
1.E+05
B1
1.E+04 B2
B3
1.E+03
BRAP
1.E+02

1.E+01

1.E+00
0 10 20 30 40 50 60 70 80 90
Temperature (°C)

Fig. 4. Isochronal curves of complex modulus G* of investigated mixtures.

Table 6 Table 7
Complex modulus values measured with DSR for each bitumen, reported Phase angle values measured with DSR for each bitumen, reported at
at temperature steps of 10 °C. temperature steps of 10 °C.
Temperature (°C) Complex Modulus |G*| (kPa) Temperature (°C) Phase angle d (°)
B0 B1 B2 B3 BRAP B0 B1 B2 B3 BRAP
0 104000 123000 110000 114000 308000 0 29 28 26 26 18
10 51600 56400 52800 56600 182000 10 38 38 35 34 23
20 16200 17200 17300 19600 74500 20 50 49 46 44 31
30 3550 3830 4200 5250 25900 30 62 61 57 54 41
40 592 658 772 1130 6790 40 72 71 67 63 52
50 103 116 138 229 1460 50 78 77 75 70 62
60 21.1 23.4 28.1 49.2 305.0 60 83 82 80 76 70
70 5.1 5.5 6.7 11.6 67.7 70 85 84 82 80 76
80 1.4 1.5 1.9 3.1 16.3 80 83 83 79 81 81
90 0.5 0.5 0.6 1.0 4.1 90 72 74 66 73 80

Fig. 5. Plot of phase angle represented as tgd versus test temperature.

 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332 329

behaviour of the mixture is proportional to the aged binder mechanical response of the bitumen is predominantly
content. When the aged binder content is equal or below elastic.
10%, there is a limited influence of the aged binder on rhe- On the basis of the previous considerations, the plot of
ological properties of the final blend. The effects of the tgd (Fig. 5) can be divided into three regions. A first region
presence of aged binder into the composite blend become (low temperatures), corresponding to a temperature range
more relevant as the RAP content raises to or exceeds of 0 °C to cross-over temperature, is where the mechanical
20%. The curves of tgd for B2 and B3 shift towards the response of bitumens is predominantly elastic. A second
BRAP curve and lower values of tgd, indicating a more elas- region (intermediate temperatures), from the cross-over
tic behaviour of the blends. temperature to approximately 70/80 °C depending on
G0 describes the amount of energy stored and elasti- RAP binder content, is where the mechanical response is
cally released during each oscillation of the DSR, while predominantly viscous and tends to become more viscous
G00 describes the energy dissipation associated with vis- as temperature increases. A third region (high
cous effects [22]. When the stored energy equals the dis- temperatures), corresponding to a temperature range of
sipated energy, the mixture experiences a viscoelastic 70/80 °C to 90 °C (end of test), is where the mechanical
transition to a predominantly elastic or viscous beha- response is predominantly viscous but tends to become
viour. The equality of G0 and G00 occurs as tgd = 1 at more elastic.
the cross-over temperature. The cross over temperature This behaviour at high temperatures can be also
allows to determine the temperature range where the observed plotting the isochronal curves of storage and loss
mechanical response of the bituminous blend is predom- modulus, reported in Figs. 6 and 7. The curves of the stor-
inantly elastic or viscous. The calculated cross-over tem- age modulus change trend at high temperatures, with ten-
peratures are reported in Table 8. dency of G0 to be horizontal, whereas the curves of the
Results indicate that the cross-over temperature loss modulus continue decreasing over the entire tempera-
increases as the content of RAP binder into the blends ture range. The plot of G0 and G00 versus temperature indi-
increases. Therefore, an increase in the aged binder content cates an increased elastic behaviour of the bitumen at high
leads to a widening of the temperature range where the temperatures, as results from the trend of tgd at high
temperatures.
Table 8
Results of cross-over temperatures and corresponding complex moduli G* 3.3.3. Black diagram
of investigated mixtures. A Black diagram is a graph of the norm of the complex
Bituminous Cross-over G* at cross-over modulus G* versus the phase angle d [21]. Data can be pre-
sample temperature (°C) temperature (MPa) sented in one plot, without any manipulation of raw data
B0 15.6 26.7 with Time-Temperature Superposition Principle (TTSP).
B1 16.1 27.1 Results in Fig. 8 show a shift of the curves towards
B2 18.6 19.6 lower values of the phase angle with the increase of the
B3 20.3 18.4
BRAP 33.1 16.6
RAP binder content into the bituminous blends. This is
caused by the dual effects of the increase in complex mod-

1.E+09

1.E+08

1.E+07
Storage modulus G' (Pa)

1.E+06

B0
1.E+05
B1
1.E+04 B2

1.E+03 B3
BRAP
1.E+02

1.E+01

1.E+00
0 10 20 30 40 50 60 70 80 90
Temperature (°C)

Fig. 6. Isochronal curves of storage component of complex modulus G0 .

330 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332

1.E+08

1.E+07

1.E+06
Loss modulus G'' (Pa)
1.E+05
B0
1.E+04 B1
B2
1.E+03
B3
BRAP
1.E+02

1.E+01

1.E+00
0 10 20 30 40 50 60 70 80 90
Temperature (°C)

Fig. 7. Isochronal curves of loss component of complex modulus G00 .

Fig. 8. Black diagram for investigated bitumens.

ulus and the decrease in phase angle. The first indicates a behaviour can be attributed to the rheology of the neat
hardening of the bitumen, the second indicates an increase binder.
in the elastic response of the mixtures as RAP binder con-
tent increases. 3.3.4. Cole-Cole diagram
The aged bitumen has little influence on the rheology of The Cole-Cole diagram is a plot of the loss modulus G00
the final blend when the RAP content in the asphalt mix- as a function of the storage modulus G0 . The graph pro-
ture is equal or below 10%, as can be noted from B0 and vides a means of representing the viscoelastic balance of
B1curves. The effect of RAP binder becomes relevant when the bitumen [14]. As for the Black diagram, the Cole-
the RAP content is equal or above 20%, as results from the Cole diagram does not require any information about test
shift of B2 and B3 curves on the Black space. temperature or frequency.
As previously observed referring to the plot of tgd, the When the RAP binder content increases, the curves
turning inward of the Black diagram curves reveals an reported on the Cole-Cole diagram (Fig. 9) shift towards
increased elastic response of the bitumens at high tempera- the lower right-hand corner of the plot, with a larger
tures. This is specific to B0, B1, B2 and B3. Therefore, this increase in storage modulus than loss modulus at a given
 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332 331

Fig. 9. Cole–Cole diagram for investigated bitumens.

temperature. Increasing the RAP binder content leads to a aged bitumen has little influence on the rheology final
more elastic, less viscous and stiffer rheological behaviour blend when the RAP content in the asphalt mixture is equal
of the mixes. or below 10%. The effects of RAP binder are significant
when the RAP content is equal or above 20%. In addition,
4. Conclusions the increase in the cross-over temperature with RAP con-
tent indicates a widening of the temperature range where
4.1. Summary of results the mechanical response of the bitumen is predominantly
elastic.
Findings from laboratory tests indicate that the aged
binder influences physical and rheological properties of
the final blend due to an interaction with the virgin binder. 4.2. Practical implications
Results of standard characterization tests show an
increase in softening point and viscosity at 160 °C and a The use of RAP into asphalt mixtures increases bitumen
decrease in penetration grade as the RAP content in the viscosity in proportion to the RAP content in the mix.
mixtures increases. This indicates a hardening of the com- Lowering bitumen viscosity may be required to achieve
posite blend caused by the presence of aged binder. adequate RAP mixes workability. An increase in produc-
The dynamic viscosity of bitumens was investigated as a tion and working temperatures may reduce bitumen viscos-
function of RAP content and temperature. Results show ity and increase mix workability, but might also cause the
that viscosity increases in proportion to the RAP binder composite blend to harden, with an undesirable stiffening
content. The RAP binder changes the chemistry of the effect on the bitumen and negative consequences on perfor-
bituminous blends, resulting in a more viscous and conse- mances of asphalt mixtures (brittleness). Using a softer and
quently harder composite bitumen. The presence of 10% hence less viscous neat binder is recommended, if needed,
(by mass of the mix) or less of RAP binder into the mixture when the RAP percentage incorporated in the mix is equal
does not affect the viscosity of the blend, whereas a RAP or above 20% (by weight of the mix). Viscosity blending
binder content of 20% or higher causes a significant charts can be used to determine the penetration grade of
increase in viscosity. The trend of viscosity with tempera- the neat bitumen for reaching the desired viscosity.
ture shows an increased temperature susceptibility as the The effects of RAP binder on the bituminous blend
RAP content increases. depend on the percentage of RAP incorporated in the mix-
The rheology of bituminous samples was investigated ture. Results from laboratory tests show that RAP can be
with the DSR analysing the complex modulus and phase incorporated into the investigated mixture at percentages
angle isochrones, the Black space and the Cole-Cole dia- up to 10% with no significant effects on properties of bitu-
gram. Results show a hardening and an increased elastic men. In this case, RAP can be added without the need to
behaviour of bitumens caused by the presence of RAP bin- perform any additional laboratory test on recovered bin-
der into the bituminous blends. The effects on the final der. As the RAP content is equal or above 20%, it’s highly
blend are in proportion to the aged binder content. The recommended to perform laboratory investigations on the
332 L. Noferini et al. / International Journal of Pavement Research and Technology 10 (2017) 322–332

recovered binder as per the procedure presented in this [11] R. West, J. Willis, M. Marasteanu, NCHRP 752, Project 09-46:
study. Improved Mix Design, Evaluation, and Materials Management
Practices for Hot Mix Asphalt with High Reclaimed Asphalt
Pavement Content, National Center for Asphalt Technology,
References Auburn, AL, 2013.
[12] 2008/98/EC, Directive 2008/98/EC of the European Parliament and
[1] Al-Qadi, Elseif, Carpenter, Reclaimed Asphalt Pavement – a of the Council of 19 November 2008 on waste and repealing certain
literature review, Illinois Center for Transportation, 2007. Directives, Off. J. Eur. Union (2008).
[2] B. Huang, W. Kingery and A. Zhang, Laboratory study of fatigue [13] L. Noferini, Performances and Durability of Asphalt Mixtures Made
characteristics of HMA mixtures containing RAP, in: International with Reclaimed Asphalt Pavement, 2016, Bologna.
symposium on long performing asphalt pavements, Auburn, 2004. [14] G. Airey, Rheological Characteristics of Polymer Modified and Aged
[3] A. Copeland, ‘‘Reclaimed asphalt pavement in asphalt mixtures: state Bitumens (Ph.D thesis), The University of Nottingham, 1997.
of the practice, FHWA-HRT-11-021”, 2011. [15] V. Vignali, F. Mazzotta, C. Sangiorgi, A. Simone, C. Lantieri, G.
[4] K. Hansen, D. Newcomb, Asphalt Pavement Mix Producion Survey Dondi, Rheological and 3D DEM characterization of potential
on Reclaimed Asphalt Pavement, Reclaimed Asphalt Shingles, and rutting of cold bituminous mastics, Constr. Build. Mater. 73 (2014)
Warm-Mix Asphalt Usage, National Asphalt Pavement Association, 339–349.
2011. [16] X. Lu, U. Isacsson, Effect of ageing on bitumen chemistry and
[5] M. Zaumanis, R. Mallick, Review of very high-content reclaimed rheology, Constr. Build. Mater. (2002) 15–22
asphalt use in plant-produced pavements: state of the art, Int. J. [17] F. Roberts, P. Kandhal, E. Brown, D. Lee, T. Kennedy, Hot Mix
Pavement Eng. (2015) 39–55 Asphalt Materials, Mixture Design, and Construction, second ed.,
[6] G. Dondi, F. Mazzotta, A. Simone, V. Vignali, C. Sangiorgi, C. Napa Education Foundation, 1996.
Lantieri, Evaluation of different short-term aging procedures with [18] J. Gaskin, On Bitumen Microstructure and the Effects of Crack
neat, warm and modified binders, Constr. Build. Mater. 106 (2016) Healing (Ph.D thesis), The University of Nottingham, 2013.
282–289. [19] N. Mechbal, M. Bousmina, Experimental study of interfacial slip
[7] W. Mogawer, T. Bennert, J. Daniel, Performance characteristics of effect on the rheological behavior of PS/PMMA blends, FSE J. (1988)
plant produced high RAP mixtures, Road Mater. Pavement Des. [20] L. Utracki, M. Kamal, Melt rheology of polymer blends, Polym. Eng.
(2012) 183–208 Sci. (1982) 96–114
[8] P. Kandhal, R. Mallick, Pavement Recycling Guidelines for State and [21] Eurobitume, Glossary of Rheological Terms – A Practical Summary
Local Governments, 1997. of the Most Common Concepts, Eurobitume, 2006.
[9] M. Zaumanis, R. Mallick, R. Frank, 100% recycled hot mix asphalt: a [22] B. Rahimzadeh, Linear and Non-linear Viscoelastic Behaviour of
review and analysis, Resour. Conserv. Recycl. 92 (2014) 230–245. Binders and Asphalts, The University of Nottingham, Nottingham,
[10] EAPA, Asphalt in Figures, European Asphalt Pavement Association, 2002.
2013.